Abstract

Financial markets are incomplete, thus for many households borrowing is possible only by accepting a
fnancial contract that specifes a fxed repayment. However, the future income that will repay this debt
is uncertain, so risk can be ineffciently distributed. This paper argues that a monetary policy of nominal
GDP targeting can improve the functioning of incomplete fnancial markets when incomplete contracts
are written in terms of money. By insulating households nominal incomes from aggregate real shocks,
this policy effectively completes fnancial markets by stabilizing the ratio of debt to income. The paper
argues the objective of replicating complete fnancial markets should receive substantial weight even in
an environment with other frictions that have been used to justify a policy of strict infation targeting.
Debt and Incomplete Financial Markets:
A Case for Nominal GDP Targeting
Kevin D. Sheedy, London School of Economics
Final conference draft
Brookings Panel on Economic Activity
March 2021, 2014
Debt and Incomplete Financial Markets:
A Case for Nominal GDP Targeting

Kevin D. Sheedy

London School of Economics

First draft: 7
th
February 2012
This version: 10
th
March 2014
Abstract
Financial markets are incomplete, thus for many households borrowing is possible only by
accepting a nancial contract that species a xed repayment. However, the future income that
will repay this debt is uncertain, so risk can be ineciently distributed. This paper argues that a
monetary policy of nominal GDP targeting can improve the functioning of incomplete nancial
markets when incomplete contracts are written in terms of money. By insulating households
nominal incomes from aggregate real shocks, this policy eectively completes nancial markets
by stabilizing the ratio of debt to income. The paper argues the objective of replicating
complete nancial markets should receive substantial weight even in an environment with
other frictions that have been used to justify a policy of strict ination targeting.
JEL classifications: E21; E31; E44; E52.
Keywords: incomplete markets; heterogeneous agents; risk sharing; nominal GDP targeting.

I thank Carlos Carvalho, Wouter den Haan, Monique Ebell, Cosmin Ilut, Albert Marcet, Matthias
Paustian, David Romer, and George Selgin for helpful comments. The paper has also beneted from the
comments of seminar participants at Banque de France, U. Cambridge, CERGE-EI,

Ecole Polytechnique,
U. Lausanne, U. Maryland, National Bank of Serbia, New York Fed, U. Oxford, PUCRio, Sao Paulo School
of Economics, U. Southampton, U. St. Andrews, U. Warwick, the Anglo-French-Italian Macroeconomics
workshop, Birmingham Econometrics and Macroeconomics conference, Centre for Economic Performance
annual conference, Econometric Society North American summer meeting, EEA annual congress, ESSET,
ESSIM, Joint French Macro workshop, LACEA, LBS-CEPR conference Developments in Macroeconomics
and Finance, London Macroeconomics workshop, Midwest Macro Meeting, and NBER Summer Institute
in Monetary Economics.

LSE, CEP, CEPR, and CfM. Address: Department of Economics, London School of Economics and
Political Science, Houghton Street, London, WC2A 2AE, UK. Tel: +44 207 107 5022, Fax: +44 207 955
6592, Email: k.d.sheedy@lse.ac.uk, Website: http://personal.lse.ac.uk/sheedy.
1 Introduction
At the heart of any argument for a monetary policy strategy lies a view of what are the most
important frictions or market failures that monetary policy should seek to mitigate. The canonical
justication for ination targeting as optimal monetary policy rests on the argument that pricing
frictions in goods markets are of particular concern (see, for example, Woodford, 2003). With
infrequent price adjustment owing to menu costs or other nominal rigidities, high or volatile ination
leads to relative price distortions that impair the ecient operation of markets, and which directly
consumes time and resources in the process of setting prices. Ination targeting is the appropriate
policy response to such frictions because it is able to move the economy closer to, or even replicate,
what the equilibrium would be if prices were exible. In other words, ination targeting is able to
undo or partially circumvent the frictions created by nominal price stickiness.
1
This paper argues that nominal price stickiness may not be the most serious friction that mone-
tary policy has to contend with. While the use of money as a unit of account in setting infrequently
adjusted goods prices is well documented, moneys role as a unit of account in writing nancial
contracts is equally pervasive. Moreover, just as price stickiness means that nominal prices fail to
be fully state contingent, nancial contracts are typically not contingent on all possible future states
of the world, for example, debt contracts that specify xed nominal repayments. Financial contracts
might not be fully contingent for a variety of reasons, but one explanation could be that transaction
costs make it prohibitively expensive to write and enforce complicated and lengthy contracts. Many
agents, such as households, would nd it dicult to issue liabilities with state-contingent repay-
ments resembling equity or derivatives, and must instead rely on simple debt contracts if they are
to borrow. Thus, in a similar way to how menu costs can make prices sticky, transaction costs can
make nancial markets incomplete.
This paper studies the implications for optimal monetary policy of such nancial-market incom-
pleteness in the form of non-contingent nominal debt contracts.
2
The argument can be understood
in terms of which monetary policy strategy is able to undo or mitigate the adverse consequences of
nancial-market incompleteness, just as ination targeting can be understood as a means of circum-
venting the problem of nominal price stickiness. For both non-contingent nominal nancial contracts
and nominal price stickiness, it is moneys role as a unit of account that is crucial, and in both cases,
optimal monetary policy is essentially the choice of a particular nominal anchor that makes money
best perform its unit-of-account function. But in spite of this formal similarity, the optimal nominal
1
In addition to the theoretical case, the more practical merits of implementing ination targeting are discussed in
Bernanke, Laubach, Mishkin and Posen (1999).
2
It is increasingly argued that monetary policy must take account of nancial-market frictions such as collateral
constraints or spreads between internal and external nance. These are dierent from the nancial frictions empha-
sized in this paper. Starting from Bernanke, Gertler and Gilchrist (1999), there is now a substantial body of work
that integrates credit frictions of the kind found in Bernanke and Gertler (1989) or Kiyotaki and Moore (1997) into
monetary DSGE models. Recent work in this area includes Christiano, Motto and Rostagno (2010). These frictions
can magnify the eects of both shocks and monetary policy actions and make these eects more persistent. But
the existence of a quantitatively important credit channel does not in and of itself imply that optimal monetary
policy is necessarily so dierent from ination targeting unless new types of nancial shocks are introduced (Faia and
Monacelli, 2007, Carlstrom, Fuerst and Paustian, 2010, De Fiore and Tristani, 2012).
1
anchor turns out to be very dierent when the friction is nancial-market incompleteness rather
than sticky prices.
One problem of non-contingent debt contracts for risk-averse households is that when borrowing
for long periods, there will be considerable uncertainty about the future income from which xed
debt repayments must be made. The issue is not only idiosyncratic uncertainty households do not
know the future course the economy will take, which will aect their labour income. Will there be a
productivity slowdown, a deep and long-lasting recession, or even a lost decade of poor economic
performance to come? Or will unforeseen technological developments or terms-of-trade movements
boost future incomes, and good economic management successfully steer the economy on a path
of steady growth? Borrowers do not know what aggregate shocks are to come, but must x their
contractual repayments prior to this information being revealed.
The simplicity of non-contingent debt contracts can be seen as coming at the price of bundling
together two fundamentally dierent transfers: a transfer of consumption from the future to the
present for borrowers, but also a transfer of aggregate risk to borrowers. The future consumption
of borrowers is paid for from the dierence between their uncertain future incomes and their xed
debt repayments. The more debt they have, the more their future income is eectively leveraged,
leading to greater consumption risk. The ip-side of borrowers leverage is that savers are able to
hold a risk-free asset, reducing their consumption risk.
To see the sense in which this bundling together of borrowing and a transfer of risk is inecient,
consider what would happen in complete nancial markets. Individuals would buy or sell state-
contingent bonds (Arrow-Debreu securities) that make payos conditional on particular states of
the world (or equivalently, write loan contracts with dierent repayments across all states of the
world). Risk-averse borrowers would want to sell relatively few bonds paying o in future states of
the world where GDP and thus incomes are low, and sell relatively more in good states of the world.
As a result, contingent bonds paying o in bad states would be relatively expensive and those paying
o in good states relatively cheap. These price dierences would entice savers to shift away from
non-contingent bonds and take on more risk in their portfolios. Given that the economy has no risk-
free technology for transferring goods over time, and as aggregate risk cannot be diversied away,
the ecient outcome is for risk-averse households to share aggregate risk, and complete markets
allow this to be unbundled from decisions about how much to borrow or save.
The ecient nancial contract between risk-averse borrowers and savers in an economy subject
to aggregate income risk (abstracting from idiosyncratic risk) turns out to have a close resemblance
to an equity share in GDP. In other words, borrowers repayments should fall during recessions
and rise during booms. This means the ratio of debt liabilities to GDP should be more stable than
it would be in a world of incomplete nancial markets where debt liabilities are xed in value while
GDP uctuates.
With incomplete nancial markets, monetary policy has a role to play in mitigating ineciencies
because private debt contracts are typically denominated in terms of money. Hence, the real degree
of state-contingency in nancial contracts is endogenous to monetary policy. If incomplete markets
were the only source of ineciency in the economy then the optimal monetary policy would aim
2
to make nominally non-contingent debt contracts mimic through variation in the real value of the
monetary unit of account the ecient nancial contract that would be chosen with complete nancial
markets.
Given that the ecient nancial contract between borrowers and savers resembles an equity
share in GDP, it follows that a goal of monetary policy should be to stabilize the ratio of debt
liabilities to GDP. With non-contingent nominal debt liabilities, this can be achieved by having a
non-contingent level of nominal income, in other words, a monetary policy that targets nominal
GDP. Nominal income thus replaces nominal goods prices as the optimal nominal anchor. The
intuition is that while the central bank cannot eliminate uncertainty about future real GDP, it can
in principle make the level of future nominal GDP (and hence the nominal income of an average
household) perfectly predictable. Removing uncertainty about future nominal income thus alleviates
the problem of nominal debt repayments being non-contingent.
A policy of nominal GDP targeting is generally in conict with ination targeting because any
uctuations in real GDP would lead to uctuations in ination of the same size and in the opposite
direction. Recessions would feature higher ination and booms would feature lower ination, or
even deation. These ination uctuations can be helpful because they induce variation in the real
value of the monetary unit of account, making it and the non-contingent debt contracts expressed
in terms of it behave more like equity. This promotes ecient risk sharing. A policy of strict
ination targeting would x the real value of the monetary unit of account, converting nominally
non-contingent debt into real non-contingent debt, which would imply an uneven and generally
inecient distribution of risk.
The ination uctuations that occur with nominal GDP targeting would entail relative-price
distortions if goods prices were sticky, so the benet of ecient risk sharing is most likely not
achieved without some cost.
3
It is ultimately a quantitative question whether the ineciency caused
by incomplete nancial markets is more important than the ineciency caused by relative-price
distortions, and thus whether nominal GDP targeting is preferable to ination targeting.
This paper presents a model that allows optimal monetary policy to be studied analytically in
an incomplete-markets economy with heterogeneous households and aggregate risk, which can be
straightforwardly calibrated for quantitative analysis. The model contains two types of households,
relatively impatient households who will choose to become borrowers, and relatively patient house-
holds who will choose to become savers. Although households dier in their time preferences, they
are all risk averse, and are all exposed to the same labour income risk. Real GDP is uncertain
because of aggregate productivity shocks, but there are no idiosyncratic shocks. The economy is
3
The model features both a trade-o between eciency in goods markets and eciency in nancial markets, and a
trade-o between the volatility of ination and the volatility of nancial-market variables. Such trade-os are implicit
in recent debates, though there is no widely accepted argument for why stabilizing prices in goods markets causes
nancial markets to malfunction. White (2009b) and Christiano, Ilut, Motto and Rostagno (2010) argue that stable
ination is no guarantee of nancial stability, and may even create conditions for nancial instability. Contrary
to these arguments, the conventional view that monetary policy should not react to asset prices is advocated in
Bernanke and Gertler (2001). Woodford (2011) makes the point that exible ination targeting can be adapted to
accommodate nancial stability concerns, and that it would be unwise to discard ination targetings role in providing
a clear nominal anchor.
3
assumed to have no investment or storage technology, and is closed to international trade. There
are no government bonds and no at money, and no taxes or scal transfers. In this world, patient
households defer consumption by lending to impatient households, who can thus bring forwards
consumption by borrowing. It is assumed the only nancial contract available is a non-contingent
nominal bond. The basic model contains no other frictions, initially assuming prices and wages are
fully exible.
The concept of a natural debt-to-GDP ratio provides a useful benchmark for monetary policy
analysis. This is dened as the ratio of (state-contingent) gross debt liabilities to GDP that would
prevail were nancial markets complete. This object is independent of monetary policy. The actual
debt-to-GDP ratio in an economy with incomplete markets would coincide with the natural debt-
to-GDP ratio if forecasts of future GDP were always correct ex post, but will in general uctuate
around it when the economy is hit by shocks. The natural debt-to-GDP ratio is thus analogous to
concepts such as the natural rate of unemployment and the natural rate of interest.
If all movements in real GDP growth rates are unpredictable then the natural debt-to-GDP ratio
turns out to be constant (or if utility functions are logarithmic, the ratio is constant irrespective
of the statistical properties of GDP growth). Even when the natural debt-to-GDP ratio is not
completely constant, plausible calibrations suggest it would have a low volatility relative to real
GDP itself.
Since the equilibrium of an economy with complete nancial markets would be Pareto ecient in
the absence of other frictions, the natural debt-to-GDP ratio also has desirable welfare properties.
A goal of monetary policy in an incomplete-markets economy is therefore to close the debt gap,
dened as the dierence between the actual and natural debt-to-GDP ratios. It is shown that doing
so eectively completes the market in the sense that the equilibrium with incomplete markets would
then coincide with the hypothetical complete-markets equilibrium. Monetary policy can aect the
actual debt-to-GDP ratio and thus the debt gap because that ratio is equal to nominal debt liabilities
(which are non-contingent with incomplete markets) divided by nominal GDP, where the latter is
under the control of monetary policy.
When the natural debt-to-GDP ratio is constant, or when the maturity of debt contracts is
suciently long, closing the debt gap can be achieved by adopting a xed target for the growth rate
(or level) of nominal GDP. With this logic, the central bank uses nominal GDP as an intermediate
target that achieves its ultimate goal of closing the debt gap. This turns out to be preferable to
targeting the debt-to-GDP ratio directly because a monetary policy that targets only a real nancial
variable would leave the economy without a nominal anchor. Nominal GDP targeting uniquely pins
down the nominal value of real incomes and thus provides the economy with a well-dened nominal
anchor.
It is important to note that in an incomplete-markets economy hit by shocks, whatever action
a central bank takes or fails to take will have distributional consequences. Ex post, there will be
winners and losers from both the shocks themselves and the policy responses. Creditors lose out
when ination is unexpectedly high, while debtors suer when ination is unexpectedly low. It might
then be thought surprising that ination uctuations would ever be desirable. However, the ination
4
uctuations implied by a nominal GDP target are not arbitrary uctuations they are perfectly
correlated with the real GDP uctuations that are the ultimate source of uncertainty in the economy,
and which themselves have distributional consequences when households are heterogeneous. For
households to share risk, it must be possible to make transfers ex post that act as insurance from
an ex-ante perspective. The result of the paper is that ex-ante ecient insurance requires ination
uctuations that are negatively correlated with real GDP (a countercyclical price level) to generate
the appropriate ex-post transfers between debtors and creditors.
It might be objected that there are innitely many state-contingent consumption allocations
which would equally well satisfy the criterion of ex-ante eciency. However, only one of these
the hypothetical complete-markets equilibrium associated with the natural debt-to-GDP ratio
could ever be implemented through monetary policy. Thus for a policymaker solely interested in
promoting eciency, there is a unique optimal policy that does not require any explicit distributional
preferences to be introduced.
The model also makes predictions about how dierent monetary policies will aect the volatility
of nancial-market variables such as the debt-to-GDP ratio. It is shown that policies implying an
inecient distribution of risk, for example, ination targeting, are associated with near-random
walk uctuations in the debt-to-GDP ratio. On the other hand, with complete nancial markets,
the persistence of uctuations in the debt-to-GDP ratio would be bounded by the persistence of
shocks to real GDP growth. When a monetary policy is adopted that allows the economy to mimic
complete nancial markets, the actual debt-to-GDP ratio inherits these less persistent dynamics.
In a model with both nominal price rigidities and incomplete nancial markets, these ndings
allow the tension between relative-price distortions and ecient risk sharing to be seen in more
familiar terms as a trade-o between price stability and nancial stability. Determining which of
these objectives is the more important in practice can be done by studying a quantitative version
of the model. Nominal price rigidity is introduced using the standard Calvo model of staggered
price adjustment. With both incomplete nancial markets and sticky prices, optimal monetary
policy is a convex combination of the optimal monetary policies that are appropriate for each of the
two frictions in isolation. After calibrating all the parameters of the model, the conclusion is that
replicating complete nancial markets should receive around 90% of the weight.
This paper is related to a number of areas of the literature on monetary policy and nancial
markets. First, there is the empirical work of Bach and Stephenson (1974), Cukierman, Lennan
and Papadia (1985), and more recently, Doepke and Schneider (2006), who document the eects of
ination in redistributing wealth between debtors and creditors. The novelty here is in studying the
implications for optimal monetary policy in an environment where ination uctuations with such
distributional eects may actually be desirable precisely because of the incompleteness of nancial
markets.
The basic idea of this paper (though not the modelling or quantitative analysis) has many
precedents in the literature. Selgin (1997) describes the ex-ante eciency advantages of falling
prices in good times and rising prices in bad times when nancial contracts are non-contingent,
5
though there is no formal modelling of the argument.
4
A survey reviewing the long history of
this idea in monetary economics is given in Selgin (1995). In recent work, Koenig (2013) advances
the risk-sharing argument for nominal GDP targeting in the context of a two-period model.
5
An
earlier theoretical paper is Pescatori (2007), who studies optimal monetary policy in an economy
with rich and poor households, in the sense of there being an exogenously specied distribution
of assets among otherwise identical households. In that environment, both ination and interest
rate uctuations have redistributional eects on rich and poor households, and the central bank
optimally chooses the mix between them (there is a need to change interest rates because prices
are sticky, with deviations from the natural rate of interest leading to undesirable uctuations in
output). A related paper is Lee (2010), who develops a model where heterogeneous households
choose less than complete consumption insurance because of the presence of convex transaction
costs in accessing nancial markets. Ination uctuations expose households to idiosyncratic labour-
income risk because households work in specic sectors of the economy, and sectoral relative prices
are distorted by ination when prices are sticky. This leads optimal monetary policy to put more
weight on stabilizing ination. Dierently from those two papers, the argument here is that ination
uctuations can actually play a positive role in completing otherwise incomplete nancial markets.
6
The idea that ination uctuations may have a positive role to play when nancial markets are
incomplete is now long-established in the literature on government debt (and has also been recently
applied by Allen, Carletti and Gale (2011) in the context of the real value of the liquidity available
to the banking system). Bohn (1988) developed the theory that non-contingent nominal government
debt can be desirable because when combined with a suitable monetary policy, ination can change
the real value of the debt in response to scal shocks that would otherwise require uctuations in
distortionary tax rates.
7
Quantitative analysis of optimal monetary policy of this kind was developed in Chari, Christiano
and Kehoe (1991) and expanded further in Chari and Kehoe (1999). One nding was that ination
needs to be extremely volatile to complete nancial markets. As a result, Schmitt-Grohe and
Uribe (2004) and Siu (2004) argued that once some nominal price rigidity is considered so that
4
Persson and Svensson (1989) is an early example of a model in the context of an international portfolio
allocation problem where it is important how monetary policy aects the risk characteristics of nominal debt.
5
Hoelle and Peiris (2013) study the eciency properties of nominal GDP targeting in a large open economy
with exible prices, and explore the question of implementability through the central banks balance sheet. Osorio-
Rodrguez (2013) examines whether there remains a role for monetary policy in completing nancial markets when
nominal debt is denominated in terms of foreign currency.
6
In other related work on incomplete markets and monetary policy, Akyol (2004) analyses optimal monetary policy
in an incomplete-markets economy where households hold at money for self insurance against idiosyncratic shocks.
Kryvtsov, Shukayev and Ueberfeldt (2011) study an overlapping generations model with at money where monetary
policy can improve upon the suboptimal level of saving by varying the expected ination rate and thus the returns
to holding money.
7
There is also a literature that emphasizes the impact of monetary policy on the nancial positions of rms or
entrepreneurs in an economy with incomplete nancial markets. De Fiore, Teles and Tristani (2011) study a exible-
price economy where there is a costly state verication problem for entrepreneurs who issue short-term nominal bonds.
Andres, Arce and Thomas (2010) consider entrepreneurs facing a binding collateral constraint who issue short-term
nominal bonds with an endogenously determined interest rate spread. Vlieghe (2010) also has entrepreneurs facing
a collateral constraint, and even though they issue real bonds, monetary policy still has real eects on the wealth
distribution because prices are sticky, so incomes are endogenous. In these papers, the wealth distribution matters
because of its eects on the ability of entrepreneurs to nance their operations.
6
ination uctuations have a cost, the optimal policy becomes very close to strict ination targeting.
This paper shares the focus of that literature on using ination uctuations to complete markets,
but comes to a dierent conclusion regarding the magnitude of the required ination uctuations
and whether the costs of those uctuations outweigh the benets. First, the benets of completing
markets in this paper are linked to the degree of risk aversion and the degree of heterogeneity among
households, which are in general unrelated to the benets of avoiding uctuations in distortionary
tax rates, and which prove to be large in the calibrated model. Second, the earlier results assumed
government debt with a very short maturity. With longer maturity debt (household debt in this
paper), the costs of the ination uctuations needed to complete the market are much reduced.
8
This paper is also related to the literature on household debt. Iacoviello (2005) examines the
consequences of household borrowing constraints in a DSGE model, while Guerrieri and Lorenzoni
(2011) and Eggertsson and Krugman (2012) study how a tightening of borrowing constraints for
indebted households can push the economy into a liquidity trap. Dierently from those papers, the
focus here is on the implications of household debt for optimal monetary policy. Furthermore, the
nding here that the presence of household debt substantially changes optimal monetary policy does
not depend on there being borrowing constraints, or even the feedback eects from debt to aggregate
output stressed in those papers. C urdia and Woodford (2009) also study optimal monetary policy in
an economy with household borrowing and saving, but the focus there is on spreads between interest
rates for borrowers and savers, while their model assumes an insurance facility that rules out the
risk-sharing considerations studied here. Finally, the paper is related to the literature on nominal
GDP targeting (Meade, 1978, Bean, 1983, Hall and Mankiw, 1994, and more recently, Sumner, 2012)
but proposes a dierent argument in favour of that policy.
The plan of the paper is as follows. Section 2 sets out a basic model and examines what monetary
policies can support risk sharing when nancial markets are incomplete. Section 3 introduces a
DSGE model that includes both incomplete nancial markets and sticky prices, and hence a trade-o
between mitigating the incompleteness of nancial markets and avoiding relative-price distortions.
Optimal monetary policy subject to this trade-o is studied in section 4. Section 5 shows how
the full model can be calibrated and presents a quantitative analysis of optimal monetary policy.
Finally, section 6 draws some conclusions.
2 A model of a pure credit economy
The analysis begins with a simplied model that studies household borrowing and saving in a nite-
horizon endowment economy with incomplete nancial markets. A full dynamic stochastic general
equilibrium model with incomplete markets together with labour supply, production, and sticky
prices is presented in section 3.
8
This point is made by Lustig, Sleet and Yeltekin (2008) in the context of government debt.
7
2.1 Assumptions
The economy contains two groups of households, borrowers (b) and savers (s), each making up
50% of a measure-one population. Household types are indexed by {b, s}. There are three time
periods t {0, 1, 2}. All households have preferences represented by the utility function:
U

= E
_
C
1
,0
1
+
C
1
,1
1
+
2
C
1
,2
1
_
, [2.1]
where C
,t
is per-person consumption by households of type at time t. All households have the
same subjective discount factor and the same coecient of relative risk aversion .
Real GDP Y
t
is an exogenous endowment. The level of GDP in period 0 is non-stochastic,
but subsequent real GDP growth rates g
t
= (Y
t
Y
t1
)/Y
t1
are uncertain. Household types are
distinguished by the shares they receive of this endowment at dierent dates The ratio of borrowers
per-person incomes to per-person real GDP is denoted by the parameter
t
, and hence household
incomes at time t are:
Y
b,t
=
t
Y
t
, and Y
s,t
= (2
t
)Y
t
. [2.2]
The income shares
t
are known with certainty in period 0. Given that both household types have
the same time preferences, the households labelled as borrowers will indeed to choose to borrow
from the savers in equilibrium when the sequence {
0
,
1
,
2
} is increasing. In other words,
borrowers are those households with initially low incomes that anticipate higher incomes in the
future, while savers have initially high incomes that they anticipate will fall in the future. In what
follows, the analysis is simplied by assuming the particular monotonic sequence of income shares
below:

0
= 1
2
E
_
(1 +g
1
)
1
(1 +g
2
)
1

,
1
= 1, and
2
= 2, [2.3]
where it is assumed that the subjective discount factor is suciently low relative to expected real
GDP growth so that
0
0.
Financial markets are incomplete in the sense that only nominal bonds can be issued or held
by households. It is assumed that borrowers cannot issue liabilities with state-contingent nominal
payos, and all non-contingent bonds are denominated in terms of money. There is assumed to be
a single type of bond traded at each date. Each bond issued in period 0 is a promise to repay 1 unit
of money in period 1, and units of money in period 2. The parameter determines the duration,
or average maturity, of the bonds ( = 0 is one-period debt, while larger values of represent
long-term debt contracts). Each bond issued in period 1 is simply a promise to repay one unit of
money in period 2. Note that in period 1, outstanding bonds from period 0 are equivalent to
newly issued bonds (old bonds are therefore counted in terms of new-bond equivalents from period
1 onwards). Households can take positive or negative positions in bonds (save or borrow) with no
limit on borrowing except being able to repay in all states of the world. There is no default, and so
all bonds are risk free in nominal terms.
Households begin with no initial assets or debts, and must leave no debts at the end of period
8
2. The net bond position per person of type- households at the end of period t is denoted by B
,t
,
the nominal bond price is Q
t
at date t, and the price of goods in terms of money is P
t
. The ow
budget identities are:
C
,0
+
Q
0
B
,0
P
0
= Y
,0
, C
,1
+
Q
1
B
,1
P
1
= Y
,1
+
(1 +Q
1
)B
,0
P
1
, and C
,2
= Y
,2
+
B
,1
P
2
, [2.4]
the term (1 +Q
1
) representing the sum of the period-1 coupon payment from period-0 bonds and
the market value of the period-2 repayment, the latter equivalent in value to newly issued bonds.
Money in this economy is simply a unit of account used in writing nancial contracts. Monetary
policy is assumed to determine the ination rate
t
= (P
t
P
t1
)/P
t1
at each date.
2.2 Equilibrium
Maximizing the utility function [2.1] subject to the budget identities [2.4] implies Euler equations
that must hold for both household types {b, s} at dates t {0, 1}:
C

,t
= E
t
_
(1 +r
t+1
)C

,t+1

. [2.5]
In equilibrium, the goods and bond markets must clear at all dates:
B
b,t
2
+
B
s,t
2
= 0; [2.6]
C
b,t
2
+
C
s,t
2
= Y
t
. [2.7]
In what follows, let c
,t
C
,t
/Y
t
denote the ratios of consumption to income, and dene the
variables d
t
, l
t
, and r
t
as follows:
d
t

1
2
(1 +Q
t
)B
b,t1
P
t
, l
t

1
2
Q
t
B
b,t
P
t
, 1 +r
t

(1 +Q
t
)P
t1
Q
t1
P
t
. [2.8]
As will be conrmed in equilibrium, B
b,t
0 and B
s,t
0, so d
t
can be interpreted as the gross
debt-to-GDP ratio (the beginning-of-period value of debt liabilities per person relative to GDP),
and l
t
as the end-of-period value of all bonds issued per person relative to GDP, referred to as the
loans-to-GDP ratio. The variable r
t
is the ex-post real return on holding bonds between periods
t 1 and t. Note that this is not the same as the interest rate on those bonds, which refers to the
ex-ante expected real return,
t
= E
t
r
t+1
. Finally, it is convenient to express the equations in terms
of the yield-to-maturity, denoted by j
t
, rather than the bond price Q
t
. Given the coupon payments
on the bonds issued in period 0 and 1, the price-yield relationships are:
Q
0
=
1
1 +j
0
+

(1 +j
0
)
2
, and Q
1
=
1
1 +j
1
. [2.9]
With the denitions [2.8], the equations of the model are:
d
t
=
_
1 +r
t
1 +g
t
_
l
t1
; [2.10a]
c
b,t
=
t
2(d
t
l
t
), c
s,t
= (2
t
) + 2(d
t
l
t
), with d
0
= 0, l
2
= 0; [2.10b]
c

,t
= E
t
_
(1 +r
t+1
)(1 +g
t+1
)

,t+1

, [2.10c]
9
where [2.10a] follows directly from [2.8], [2.10b] is derived from the budget identities [2.4] and market
clearing condition [2.6], and [2.10c] from the Euler equations [2.5]. Finally, [2.8] and the denition
of the yield-to-maturity in [2.9] imply:
1 +r
1
=
_
1 +j
0
1 +
1
_
_
1 +

1+j
1
1 +

1+j
0
_
, and 1 +r
2
=
1 +j
1
1 +
2
. [2.11]
It is assumed that the parameter restriction in [2.3] always holds in what follows.
In the case where there is no uncertainty about the path of real GDP (g
1
= g
1
and g
2
= g
2
, where
g
1
and g
2
are non-stochastic), and where there are no unexpected changes in ination (
1
=
1
and

2
=
2
), the system of equations [2.10a][2.10c] and [2.11] has the following solution:
c
b,t
= c
s,t
= 1, 1 + r
1
=
(1 + g
1
)

, and 1 + r
2
=
(1 + g
2
)

; [2.12a]

d
1
=

2
(1 + g
2
)
1
, and

d
2
=
1
2
; [2.12b]

l
0
=

2
2
(1 + g
1
)
1
(1 + g
2
)
1
, and

l
1
=

2
(1 + g
2
)
1
. [2.12c]
The equilibrium interest rates (equal here to the ex-post real returns r
1
and r
2
) are identical to
what would prevail if there were a representative household. The income shares in [2.3] imply that
in the absence of shocks, borrowers and savers would have the same levels of consumption. Given
the levels of income and consumption, the implied nal debt-to-GDP ratio is 50%, and given [2.3],
the debt-to-GDP and loans-to-GDP ratios at earlier dates are discounted values of the nal debt-
to-GDP ratio (adjusted for any real GDP growth). This steady state is independent of monetary
policy. The values of
1
and
2
together with the real interest rates r
1
and r
2
determine the nominal
bond yields

j
0
and

j
1
.
2.3 The complete nancial markets benchmark
Consider an hypothetical economy that has complete nancial markets but is otherwise identical
to the economy described above. Households now have access to a complete set of state-contingent
bonds (traded sequentially, period-by-period), denominated in real terms without loss of generality.
Let F

,t
denote the net portfolio of contingent bonds per person held between periods t and t +1 by
households of type (the asterisk signies complete nancial markets). The prices of these securities
in real terms relative to the conditional probabilities of the states at time t are denoted by K
t+1
, so
E
t
[K
t+1
F

,t+1
] is the date-t cost of the date-t + 1 payo F

,t+1
.
In this version of the model, the ow budget identities [2.4] are replaced by:
C

,t
+E
t
[K
t+1
F

,t+1
] = Y
,t
+F

,t
, [2.13]
together with initial and terminal conditions F

,0
= 0 and F

,3
= 0. The Euler equations for
maximizing utility [2.1] subject to [2.13] are:

_
C

b,t+1
C

b,t
_

= K
t+1
=
_
C

s,t+1
C

s,t
_

, [2.14]
10
which must hold in all states of the world. The market-clearing condition F

b,t
/2+F

s,t
/2 = 0 replaces
equation [2.6].
To relate the economy with complete markets to its incomplete-markets equivalent, consider the
following denitions of variables d

t
, l

t
, and r

t
that will be seen to be the equivalents of the debt-to-
GDP ratio d
t
, the loans-to-GDP ratio l
t
, and the ex-post real return r
t
in the incomplete-markets
economy (as given in [2.8]):
d

t

1
2
F

b,t
Y
t
, l

t

1
2
E
t
[K
t+1
F

b,t+1
]
Y
t
, and 1 +r

t

F

b,t
E
t1
[K
t
F

b,t
]
. [2.15]
Debt in an economy with complete nancial markets refers to the total gross value of the contingent
bonds repayable in the realized state of the world. Loans refers to the value of the whole portfolio
of contingent bonds issued by borrowers, and the (gross) ex-post real return is the state-contingent
value of the bonds repayable relative to the value of all the bonds previously issued.
The denitions [2.15] directly imply that equation [2.10a] must hold in terms of c

,t
, d

t
, l

t
,
and r

t
. The budget identities [2.13] and the contingent bond-market clearing conditions imply
that [2.10b] holds in terms of the variables dened in [2.15]. Since E
t
[(1 + r

t+1
)K
t+1
] = 1 follows
from the denition in [2.15], the rst-order conditions [2.10c] imply that [2.10c] must hold in terms
of c

,t
and r

t
. Hence, the block of equations [2.10a][2.10c] applies to both the incomplete- and
complete-markets economies.
The rst-order condition [2.14] with complete markets has stronger implications than equation
[2.10c], though. It also requires
c

b,t+1
c

b,t
=
c

s,t+1
c

s,t
, [2.16]
to hold in all states of the world. This equation says that consumption growth rates must always be
equalized between borrowers and lenders, in other words, households use complete nancial markets
to share risk. This is not generally an implication of the equilibrium conditions [2.10a][2.10c] and
[2.11] with incomplete nancial markets. Furthermore, since a complete-markets economy has no
restriction on the types of assets that households can buy and sell, equation [2.11] that determines the
ex-post real return on a portfolio of nominal bonds is now irrelevant to determining the equilibrium
of the complete-markets economy. The ex-post real return is now determined implicitly by the
portfolio of contingent securities that ensures the risk-sharing condition [2.16] holds.
The complete-markets equilibrium can be obtained analytically by solving the system of equa-
tions [2.10a][2.10c] and [2.16] (again under the parameter restriction in [2.3]). The equilibrium
consumption-GDP ratios are c

,t
= 1, so there is full risk sharing between borrowers and savers,
meaning that all households consumption levels perfectly co-move in response to shocks (the con-
sumption levels are equal owing to the parameter restriction [2.3]). Complete nancial markets
therefore allocate consumption eciently across states of the world, as well as over time. The equi-
librium values of other variables have similar expressions to those found in the non-stochastic case
[2.12] except that certain outcomes are replaced by expected values conditional on earlier informa-
11
tion:
d

1
=

2
E
1
_
(1 +g
2
)
1

, and d

2
=
1
2
; [2.17a]
l

0
=

2
2
E
_
(1 +g
1
)
1
(1 +g
2
)
1

, and l

1
=

2
E
1
_
(1 +g
2
)
1

. [2.17b]
The nal debt-to-GDP ratio d

2
is non-stochastic, and earlier debt-to-GDP ratios depend only on
conditional expectations of future real GDP growth rates. Since the realization of shocks in period 1
can change these conditional expectations, d

1
and l

1
are stochastic in general. The expressions for l

0
,
l

1
, and d

1
can be interpreted as the present discounted values in periods 0 or 1 of a payo proportional
to the nal debt-to-GDP ratio d

2
, evaluated using prices of contingent securities (which are equal
to households common stochastic discount factor according to [2.14]). Intuitively, the complete-
markets portfolio is an equity share in future real GDP. This supports risk sharing by allowing the
repayments of borrowers to move exactly in line with their incomes.
Finally, observe that the complete-markets equilibrium has a particularly simple form in two
special cases. If the utility function is logarithmic ( = 1) or real GDP follows a random walk (g
t
is i.i.d.) then d

t
and l

t
are all non-stochastic:
d

1
=

2
, d

2
=
1
2
, l

0
=

2
2
, and l

1
=

2
, where = E
_
(1 +g
t
)
1

. [2.18]
Note that the complete-markets equilibrium is entirely independent of monetary policy in all cases.
2.4 Replicating complete nancial markets
The block of equations [2.10a][2.10c] is common to the equilibrium conditions irrespective of
whether nancial markets are complete or not. The only dierence between the equilibrium con-
ditions is that the incomplete-markets economy includes [2.11] instead of [2.16] in the complete-
markets economy. Since equation [2.11] includes the ination rates
1
and
2
, the ability of monetary
policy to engineer a suitable state-contingent path for ination means that the ex-post real returns
in [2.11] can be chosen to generate the same consumption allocation as implied by the risk-sharing
condition [2.16].
9
This is in turn equivalent to ensuring the actual debt-to-GDP ratio d
t
mimics its
hypothetical equilibrium value d

t
in the economy with complete nancial markets.
The monetary policy that replicates complete nancial markets in this way turns out to be
a nominal GDP target. Nominal GDP is denoted by N
t
= P
t
Y
t
, and its growth rate by n
t
=
(N
t
N
t1
)/N
t1
. Since it is assumed that monetary policy can determine a state-contingent path
for ination
t
, and as real GDP growth g
t
is exogenous, monetary policy can equally well be
specied as a sequence of nominal GDP growth rates n
t
. Equation [2.11] for the ex-post real returns
on nominal bonds can be written in terms of nominal GDP growth as follows:
1 +r
1
1 +g
1
=
_
1 +j
0
1 +n
1
_
_
1 +

1+j
1
1 +

1+j
0
_
, and
1 +r
2
1 +g
2
=
1 +j
1
1 +n
2
. [2.19]
9
That monetary policy is able exactly to replicate complete nancial markets is owing to the model having a
representative borrower and a representative saver. With heterogeneity within these groups as well as between
them, exact replication will generally not be possible.
12
Consider rst the replication argument in either of the special cases of log utility ( = 1) or
real GDP following a random walk (g
t
is an i.i.d. stochastic process). In these special cases, the
complete-markets debt-to-GDP ratios [2.18] that monetary policy is aiming to replicate are non-
stochastic. Suppose monetary policy sets a non-stochastic path for nominal GDP, that is, n
1
= n
1
and n
2
= n
2
for some constants n
1
and n
2
. If the replication is successful, c
,t
= 1, and so both
households Euler equations [2.10c] are satised when 1 = E
t
[(1 +r
t+1
)(1 +g
t+1
)

]. With n
2
= n
2
and [2.19] this requires for t = 1:
1 +j
1
1 + n
2
=
1
E
1
[(1 +g
2
)
1
]
=
1

,
using the denition of from [2.18]. It then follows from [2.19] that (1 + r
2
)/(1 + g
2
) = 1/.
From the complete-markets solution [2.18] together with [2.10a], (1 + r

2
)/(1 + g
2
) = 1/, so this
monetary policy ensures the ex-post real return r
2
on nominal bonds is identical to that on the
complete-markets portfolio r

2
for all realizations of shocks. Similarly, with n
1
= n
1
and the solution
/(1 +j
1
) = /(1 + n
2
) from above, the Euler equation at t = 0 requires:
1 +j
0
1 + n
1
1
1 +

1+j
0
=
1
1 +

1+ n
2
1

,
and hence (1 + r
1
)/(1 + g
1
) = 1/, which coincides with (1 + r

1
)/(1 + g
1
) = 1/. Therefore,
r
1
= r

1
, and r
2
= r

2
, under this monetary policy, which establishes that d
t
= d

t
and c
,t
= c

,t
. A
nominal GDP target with any non-stochastic rates of nominal GDP growth succeeds in replicating
the complete-markets equilibrium and supporting risk-sharing among borrowers and savers. The
intuition is that if the numerator of the debt-to-GDP ratio (expressed in monetary units) is xed
because nominal debt liabilities are not state contingent, the ratio can be stabilized by targeting the
denominator (expressed in monetary units), that is, ensuring that nominal incomes are predictable.
There is one other special case in which a monetary policy that makes nominal GDP growth
perfectly predictable manages to replicate complete nancial markets, even when the complete-
markets debt-to-GDP ratio d

1
is stochastic. This is the case where borrowers do not need to issue
any new debt and do not need to renance any existing debt after the initial time period. In the
model, this corresponds to the limiting case of pure long-term bonds where ( is the ratio
of the period-2 and period-1 coupon payments on a bond issued in period 0). If monetary policy
ensures that (1+n
1
)(1+n
2
) = (1+ n)
2
for some non-stochastic n then the equilibrium of the economy
will coincide with the hypothetical complete-markets equilibrium. Unlike the earlier special cases,
here it is not necessary that both the period 1 and 2 nominal GDP growth rates are non-stochastic,
only that the cumulated growth rate over both time periods is perfectly predictable. Intuitively,
with long-term debt, monetary policy needs only to ensure that nominal incomes are predictable
when debt is actually repaid.
When is nite, some existing debt must be repaid by borrowers in period 1, requiring them
to issue some new bonds if they are to continue to borrow until period 2. This exposes them to
risk coming from uncertainty about the interest rate that will prevail in period 1. A monetary
policy that aims to replicate complete nancial markets must then address renancing risk as well
13
as income risk. In general, this requires a target for nominal GDP growth that changes when shocks
occur, though as will be seen, there will still be a long-run target for nominal GDP that is invariant
to shocks. If the maturity parameter is positive and nite then the class of monetary policies
that replicate complete nancial markets are given by the following nominal GDP growth rates in
periods 1 and 2:
1 +n
1
= (1 + n
1
)
E[(1 +g
2
)
1
]
E
1
[(1 +g
2
)
1
]
, and 1 +n
2
= (1 + n
2
)
E
1
[(1 +g
2
)
1
]
E[(1 +g
2
)
1
]
, [2.20]
where n
1
and n
2
are any non-stochastic growth rates (these would be the actual nominal GDP
growth rates in the absence of shocks). Note that any such monetary policy has the implication
that
(1 +n
1
)(1 +n
2
) = (1 + n
1
)(1 + n
2
), [2.21]
so the long-run target for nominal GDP must be non-stochastic. In the special cases of = 1 or
g
t
being i.i.d. that were analysed earlier, the requirements on nominal GDP growth rates in [2.20]
reduce simply to n
1
= n
1
and n
2
= n
2
. Intuitively, the problem of renancing risk is absent in these
special cases, albeit for a dierent reason in each. When real GDP growth rates are independent
over time, there is no news that changes expected future real GDP growth, and thus no reason for
the equilibrium real interest rate to vary over time. On the other hand, with log utility, the real
interest rate changes by the same amount and in the same direction as any revision to expectations
of future growth. Higher real interest rates are then exactly oset by expectations of an improvement
in future incomes relative to current incomes. This leaves monetary policy needing only to provide
insurance against uctuations in current incomes, for which a predictable nominal GDP growth rate
suces.
As discussed above, in the special case of pure long-term debt ( ), [2.21] is needed, but
[2.20] need not hold. In the special case of pure short-term debt ( = 0), it can be shown that only
the condition on n
1
in [2.20] is needed, together with the restriction that n
2
takes on a value that
is perfectly predictable in period 1.
2.5 Consequences of optimal monetary policy for ination
Optimal monetary policies that replicate complete nancial markets have been characterized as
nominal GDP targets. By denition, stabilizing nominal GDP when there are uctuations in real
GDP entails uctuations in ination. That ination uctuates is not in itself the desirable feature
of these policies, but rather that ination displays a negative correlation with real GDP growth.
In other words, there is an optimal degree of countercyclicality of the price level. In cases where a
constant nominal GDP growth rate is optimal, ination should move in the opposite direction to
real GDP growth by the same amount.
In general, when the required nominal GDP growth rate is not independent of the realization of
14
shocks (see [2.20]), the implied ination rates are:
1 +
1
=
(1 + n
1
)
(1 +g
1
)
E[(1 +g
2
)
1
]
E
1
[(1 +g
2
)
1
]
, and 1 +
2
=
(1 + n
1
)
(1 +g
2
)
E
1
[(1 +g
2
)
1
]
E[(1 +g
2
)
1
]
. [2.22]
Now suppose that real GDP does not follow a random walk (g
t
is not i.i.d.), and utility is not
logarithmic ( = 1). Take the realistic case where > 1 (high risk aversion, low intertemporal
substitution), and where there is some mean reversion in the level of real GDP after a shock has
occurred (so the growth rates g
1
and g
2
are negatively correlated). Under these assumptions, the
term E
1
[(1 +g
2
)
1
] in [2.22] moves in the same direction as (1 +g
1
) following a shock to real GDP
growth in period 1. This means the required change in ination is even larger than what would be
implied by non-stochastic target for the nominal GDP growth rate. Intuitively, any unexpected fall
in g
1
leads to an increase in the real interest rate because g
2
is expected to rise, and the increase in
the real interest rate is larger. The complete-markets debt-to-GDP ratio d

1
from [2.17a] thus falls
because the interest rate eect dominates the growth eect. To decrease the debt-to-GDP ratio in
an economy with nominal bonds, it is necessary for ination to rise by more than the fall in.
In period 2, equation [2.22] shows that both predictable and unpredictable changes in real GDP
growth g
2
require movements in period 2 ination in the opposite direction to real GDP growth.
Therefore, even in the case where optimal monetary policy is not a perfectly predictable nominal
GDP growth rate, it is unlikely that a constant ination rate would ever be optimal. In other words,
the required nominal GDP growth rate would not ever be likely to move by more than real GDP
growth in the same direction. To reinforce this conclusion, note that [2.21] implies the cumulated
ination rate over periods 1 and 2 must be
(1 +
1
)(1 +
2
) =
(1 + n
1
)(1 + n
2
)
(1 +g
1
)(1 +g
2
)
.
This means that the price level must be countercyclical on average over periods 1 and 2, and the
only case in which the long-run price level should be predictable is unrealistic one in which the
long-run level of real GDP is perfectly predictable.
Writing a target for monetary policy in terms of ination and real GDP growth might be reminis-
cent of so-called exible ination targeting, and it might be tempting to interpret a nominal GDP
target as simply a relabelling of monetary policies of that kind. But aside from the rationale being
very dierent from that invoked to justify exible ination targeting, there is a fundamental dier-
ence between the two policies. Flexible ination targeting can be formalized as a target criterion
in both ination and the output gap, whereas here, the target criterion includes the actual growth
rate of output, not the output gap. Monetary policy is eective in completing nancial markets
precisely because the level of output that appears in the target is not adjusted for any unexpected
changes in potential output, even though these have consequences for ination.
2.6 Consequences of suboptimal monetary policy for nancial markets
Following a monetary policy that replicates complete nancial markets has consequences for uc-
tuations in ination. Similarly, following a monetary policy that stabilizes ination has observable
15
implications for uctuations in nancial markets in addition to failing to replicate the risk sharing
of complete nancial markets. This is because the behaviour of variables such as the debt-to-GDP
ratio is very dierent in a complete-markets economy than in an economy with incomplete markets
and a sub-optimal monetary policy.
In the hypothetical complete-markets economy, the equilibrium debt-to-GDP ratio has some
striking properties. Intuitively, the stock of outstanding debt relative to GDP ought to behave
like a state variable because debt liabilities are predetermined. However, with access to complete
nancial markets, households optimizing behaviour leads them to choose a portfolio of contingent
securities such that the implied debt-to-GDP ratio is actually a purely forward-looking variable: all
the expressions for d

t
and l

t
in [2.17a] depend only on expectations of future real GDP growth.
The intuition for this surprising result is that given the budget identities of households, full risk
sharing requires the nancial wealth of savers and the nancial liabilities of borrowers to move in
line with the value of future labour income. The debt-to-GDP ratio must therefore behave like an
asset price rather than a state variable: it must reect a forecast of the economys future prospects
rather than a record of past choices and shocks. It follows that in the complete-markets economy,
past realizations of shocks would have no correlation with the current debt-to-GDP ratio except to
the extent that these have predictive power for the economys future fundamentals.
When monetary policy is used to replicate complete nancial markets in an incomplete-markets
economy, uctuations of ination are used to aect the real value of nominal debt so as to mimic
the behaviour of the complete-markets debt-to-GDP ratio. But if monetary policy does not repli-
cate complete nancial markets then the debt-to-GDP behaves in line with what simple intuition
suggests: it is intrinsically serially correlated, so past shocks have a persistent eect on the subse-
quent evolution of the level of debt. With incomplete nancial markets, optimizing behaviour by
households leads to consumption smoothing over time, but not generally across dierent states of
the world. Thus, following a shock, households nancial wealth may diverge from the present dis-
counted value of future labour income. Moreover, when this divergence occurs, because households
spread out the adjustment of consumption over time, the consequences of the shock for nancial
wealth and consumption are long lasting.
10
To illustrate these claims, assume for simplicity that all uncertainty about real GDP in period
2 has been resolved by period 1. Suppose that monetary policy in period 1 (together with the
monetary policy expected in period 2) has failed to replicate complete nancial markets, so the
10
The stark dierence compared to complete nancial markets is analogous to some well-known results from the
literature on optimal scal policy under dierent assumptions about the completeness of the nancial markets a
government has access to. That literature considers an environment where the government aims to nd the least
distortionary means of nancing a stochastic sequence of government spending, given that available scal instruments
entail distortions which are convex in tax rates. With incomplete nancial markets (in the sense that the government
can issue only non-contingent bonds), Barro (1979) nds the government should aim to smooth tax rates, which
implies the stock of government debt follows a random walk. On the other hand, Lucas and Stokey (1983) assume
the government can issue a full set of contingent bonds. In that case, the government now smooths taxes across
states-of-the-world as well as time, and this means the value of outstanding government liabilities is now a purely
forward-looking variable depending on expectations of future scal fundamentals. These ndings for the behaviour
of government debt mimic the ndings for household debt here, with consumption smoothing (whether across time
or also across states of the world) playing the role of tax smoothing in the analysis of optimal scal policy.
16
actual debt-to-GDP ratio d
1
diers from its complete-markets equivalent d

1
. In period 2 there
are no new shocks to react to, and it is assumed that there are no unexpected shifts in monetary
policy unrelated to the economys fundamentals. It follows from [2.10a] and [2.10b] that there is no
uncertainty about c
,2
as of period 1, and so the Euler equations [2.10c] imply c
b,2
/c
b,1
= c
s,2
/c
s,1
.
Given [2.10b], this means c
b,2
= c
b,1
and c
s,2
= c
s,1
. Solving [2.10a][2.10b] for c
,1
, c
,2
, and d
2
in
terms of the state variable d
1
, and comparing to the complete-markets equilibrium in [2.17a]:
d
2
d

2
=
d
1
d

1
1 +(1 +g
2
)
1
, and
c
b,1
c

b,1
= c
b,2
c

b,2
=
2(d
1
d

1
)
1 +(1 +g
2
)
1
, c
s,1
c

s,1
= c
s,2
c

s,2
=
2(d
1
d

1
)
1 +(1 +g
2
)
1
.
This analysis shows that the equilibrium debt-to-GDP ratio is positively related to the previous
periods debt-to-GDP ratio when monetary policy fails to replicate complete nancial markets. The
intuition is that consumption smoothing implies that any deviation from the complete-markets
consumption allocation persists over time. While the positive serial correlation of the debt-to-GDP
seems an obvious property, it is important to remember that the debt-to-GDP ratio in a complete-
markets economy does not have this feature: it is a purely forward-looking variable.
2.7 Consequences of ination indexation of bonds
Given that savers holding nominal bonds are exposed to the risk of ination uctuations, it might be
thought desirable that bonds be indexed to ination. To analyse this question, consider an otherwise
identical economy with incomplete markets but where all bonds have coupons indexed to the price
level P
t
. To simplify the analysis, suppose all bonds have a maturity of one time period ( = 0).
Writing down ow budget constraints analogous to [2.4] and following all the subsequent steps
leads to a set of equilibrium conditions including [2.10a][2.10c], which now hold in terms of c

,t
, d

t
,
l

t
, and r

t
(the

superscript signies the values of these variables with ination indexation). The set
of equilibrium conditions diers only from the incomplete-markets economy with nominal bonds in
that equation [2.11] for the ex-post real return is replaced by
1 +r

1
= 1 +y
0
, and 1 +r

2
= 1 +y
1
, [2.23]
where y
t
is the real yield-to-maturity on the indexed bonds, which is both the ex-ante interest rate
and the ex-post real return on these bonds (

t
= y
t
= r

t+1
).
It is clear that the equilibrium for all real variables is now independent of monetary policy. A
comparison of [2.11] and [2.23] shows that the equilibrium coincides with what the equilibrium of
the nominal-bonds economy would be if monetary was strict ination targeting (with non-stochastic
ination rates
1
=
1
and
2
=
2
).
This result shows that moving to an economy with indexed bonds is generally worse if monetary
policy aims to replicate complete nancial markets. Now, whatever the central bank tries to do
has no real eects, so there is no monetary policy intervention that can complete nancial markets.
All that happens is indexation locks in the generally sub-optimal outcome that would prevail with
17
strict ination targeting. This is in spite of the fact that savers are now protected from ination
uctuations. The intuition for these ndings is best understood by considering an economy in which
both nominal and indexed bonds are available.
2.8 Portfolio choice and the ination risk premium
The analysis so far has assumed only a single type of bond is available. While the general case of
many dierent assets is beyond the scope of this paper, it is helpful to consider the consequences of
there being both nominal and indexed bonds that can be bought or issued by households. Assume
for simplicity that all bonds have a maturity of one time period ( = 0).
Writing down a ow budget constraint for this case analogous to [2.4] and following the subse-
quent steps leads to equilibrium conditions [2.10a][2.10b] as before, which now hold in terms of c

,t
,
d

t
, l

t
, and r

t
(the

superscript signies the case of both nominal and indexed bonds). The ex-post
real return r

t
is now the return on the whole portfolio of nominal and indexed bonds:
r

t
= (1 s
t1
)r
t
+s
t1
r

t
, [2.24]
where s
t
denotes the portfolio share in indexed bonds (borrowers and savers must have the same
portfolio shares in equilibrium), and the ex-post real returns r
t
and r

t
on nominal and indexed
bonds individually are given by equations [2.11] and [2.23] as before. In this version of the model,
the Euler equations [2.10c] must hold for both r
t+1
and r

t+1
for both borrowers and savers. The
ination risk premium
t
is dened as
t
= E
t
[(1 +r
t+1
)/(1 +r

t+1
)] 1.
First consider the implications of strict ination targeting (
1
=
1
and
2
=
2
). Using the Euler
equations [2.10c] and comparing equations [2.11] and [2.23], it follows immediately that r
t
= r

t
, and
hence r

t
= r
t
= r

t
for any portfolio share s
t
. The equilibrium of the economy is thus the same
as that where strict ination targeting is followed in an economy with only nominal bonds (any
portfolio share s
t
is an equilibrium). The ination risk premium is zero (
t
= 0).
Now consider the nominal GDP target [2.20] that replicates complete markets in the economy
with only nominal bonds. The policy still achieves this goal in the economy with both nominal
and indexed bonds, with the equilibrium portfolio share in indexed bonds being zero (s
t
= 0). The
average ination risk premia are given by:
E
1
=
E[(1 +g
1
)(1 +g
2
)
1
] E[(1 +g
1
)

]
E[(1 +g
1
)
1
(1 +g
2
)
1
]
1, E
2
=
E[(1 +g
2
)] E[(1 +g
2
)

]
E[(1 +g
2
)
1
]
1.
The expression for E
2
is always positive, and E
1
is generally positive (unambiguously so if = 1
or g
t
is i.i.d.).
These results show that for both strict ination targeting and a nominal GDP target that
replicates complete nancial markets, the equilibrium outcomes with both types of bonds are the
same as those of an economy with only nominal bonds. For strict ination targeting, this is simply
because the absence of ination uctuations makes both types of bond equivalent, thus any portfolio
share can be an equilibrium. Perhaps more surprisingly, the ability to hold indexed bonds does not
change the equilibrium when monetary policy pursues a policy that replicates complete nancial
18
markets, which does entail uctuations in ination. The equilibrium portfolio share of indexed
bonds is zero in this case, so savers do not attempt to protect themselves from ination risk. The
reason is the existence of an ination risk premium, which means that savers earn higher average
returns by holding nominal bonds, which compensate them for the risk they bear.
This perhaps shifts the question to why borrowers continue to issue nominal bonds when lower
real interest rates are available on indexed bonds. However, for borrowers, it is the indexed bond
that is riskier and the nominal bond that is safer because the former obliges the borrower to make
the same real repayments irrespective of real income. When monetary policy replicates complete
nancial markets, the countercyclical price level makes nominal bonds behave like equity, so coupons
have a lower real value when real incomes are low, providing insurance to borrowers, for which they
are willing to pay a higher average real interest rate. The higher average real interest rate on nominal
bonds can thus equally well be seen as an insurance premium for borrowers as a risk premium
for savers, and this ination risk premium is actually a desirable feature of monetary policy. The
ination uctuations with nominal GDP targeting are not simply generating risk for savers; ination
is a hedge for borrowers against the underlying real risk in the economy.
11
2.9 Goals and intermediate targets
The argument for a nominal GDP target in this paper is that it replicates the debt-to-GDP ratio
that would be found if the economy had complete nancial markets. Thus, the debt-to-GDP ratio
(and nally the implied consumption allocation) is the ultimate objective of policy. Nominal GDP
is simply an intermediate target which helps achieve that goal. It might then be thought preferable
to target the debt-to-GDP ratio directly if this is what monetary policy is actually seeking to
inuence. However, as [2.20] shows, there are many possible nominal GDP growth rates consistent
with replicating complete nancial markets. So an obvious pitfall of targeting the debt-to-GDP
ratio directly is that it would fail to provide the economy with a nominal anchor. Once a specic
nominal GDP target consistent with [2.20] is chosen, this policy both replicates complete markets
and provides the economy with a nominal anchor.
2.10 Eciency, fairness, and the distributional eects of policy
The earlier analysis took it for granted that replicating complete nancial markets ought to be
an objective of monetary policy. The implicit justication for this is the eciency properties of
the complete-markets equilibrium, which is Pareto ecient in the absence of any other distortions.
Thus, in spite of policy having distributional eects, the optimal policy can be viewed as supporting
11
It would also be possible to replicate complete nancial markets in the two-bond economy with a lower variance
of ination as long as monetary policy deviates from strict ination targeting, ensuring there is some correlation
between ination and real GDP growth. The reduction of the covariance of ination and real GDP growth would
require all households to hold a positive or negative position in both nominal and indexed bonds, with the size of the
gross positions increasing as the covariance between ination and real GDP growth shrinks. However, since the gross
positions are larger, monetary policy errors (not generating exactly the covariance between ination and real growth
that households were expecting) lead to larger deviations from complete nancial markets than when households only
need to buy or sell one type of bond. These issues are left for future research.
19
(ex-ante) Pareto eciency in much the same way that analyses of optimal monetary policy have
pointed to other ineciencies that policy might correct. However, this justication is incomplete in
one important respect: while the complete-markets equilibrium is Pareto ecient, there are innitely
many other consumption allocations that would equally well satisfy the criterion of Pareto eciency.
What might support singling out one particular Pareto-ecient allocation as the target for policy
when the choice among this set necessarily entails taking a stance on distributional questions?
Consider a policymaker who evaluates welfare using a weighted sum of the utilities of borrowers
and savers:
W =

b
2
U
b
+

s
2
U
s
, [2.25]
where

is the weight assigned to each household of type {b, s}, which reect the policymakers
distributional preferences. Suppose the policymaker were a social planner who has access to a full
set of state-contingent transfers between households. In this case, the optimal monetary would be
to maximize the welfare function [2.25] subject only to the economys resource constraint, which is
the goods-market clearing condition [2.7]. The rst-order conditions for this maximization problem
are:

,t
=
t
, [2.26]
where
t
denotes Lagrangian multiplier on resource constraint at time t. This rst-order condition
implies the risk-sharing condition [2.16] must hold for any Pareto weights

. The resulting con-

sumption allocation would be one of many rst-best (Pareto ecient) allocations. Conversely, any
Pareto ecient allocation is a solution of the planners problem for some Pareto weights. The set of
rst-best consumption allocations is characterized by those that satisfy the resource constraint and
the risk-sharing condition. The equilibrium of the economy with complete nancial markets is one
of these rst-best allocations.
Unlike the hypothetical social planner, a policymaker setting monetary policy can only inuence
the allocation of consumption through monetary policys impact on ination and interest rates.
Given the distributional preferences of the central bank as represented by

, optimal monetary
policy would maximize welfare [2.25] subject to all the equilibrium conditions [2.10a][2.10c] and
[2.11] of the economy with incomplete nancial markets. This maximization problem is subject
to additional constraints beyond the resource constraint, so the resulting second-best consumption
allocation is generally not Pareto ecient because there is a trade-o between eciency and the
distributional preferences of the policymaker.
However, the nominal GDP targets described earlier are known to replicate the complete-markets
equilibrium, so monetary policy can achieve at least one rst-best allocation. The complete-markets
equilibrium is also the only rst-best allocation implementable through monetary policy. The in-
tuition for this result is that the risk-sharing condition is necessary for any allocation to be rst
best. It is also the only equation that distinguishes the equilibrium conditions with complete markets
from the equilibrium conditions with incomplete markets, with the latter being the implementability
constraints on monetary policy.
20
Therefore, an implicit assumption that justies the focus on replicating the complete-markets
equilibrium through nominal GDP targeting is that the policymaker has a lexicographic preference
for eciency over any explicit distributional concerns: any rst-best allocation is always favoured
over a non-rst-best allocation. In determining optimal monetary policy, the central bank sim-
ply adopts whatever Pareto weights are associated with the only implementable ex-ante ecient
allocation.
2.11 Discussion
The importance of the arguments for nominal GDP targeting in this paper obviously depends on
the plausibility of the incomplete-markets assumption in the context of household borrowing and
saving. It seems reasonable to suppose that households will not nd it easy to borrow by issuing
Arrow-Debreu state-contingent bonds, but might there be other ways of reaching the same goal?
Issuance of state-contingent bonds is equivalent to households agreeing loan contracts with nancial
intermediaries that specify a complete menu of state-contingent repayments. But such contracts
would be much more time consuming to write, harder to understand, and more complicated to en-
force than conventional non-contingent loan contracts, as well as making monitoring and assessment
of default risk a more elaborate exercise.
12
Moreover, unlike rms, households cannot issue securities
such as equity that feature state-contingent payments but do not require a complete description of
the schedule of payments in advance.
13
Another possibility is that even if households are restricted to non-contingent borrowing, they can
hedge their exposure to future income risk by purchasing an asset with returns that are negatively
correlated with GDP. But there are several pitfalls to this. First, it may not be clear which asset
reliably has a negative correlation with GDP (even if GDP securities of the type proposed by
Shiller (1993) were available, borrowers would need a short position in these). Second, the required
gross positions for hedging may be very large. Third, a household already intending to borrow will
need to borrow even more to buy the asset for hedging purposes, and the amount of borrowing may
be limited by an initial down-payment constraint and subsequent margin calls. In practice, a typical
borrower does not have a signicant portfolio of assets except for a house, and housing returns most
likely lack the negative correlation with GDP required for hedging the relevant risks.
In spite of these diculties, it might be argued the case for the incomplete markets assumption
is overstated because the possibilities of renegotiation, default, and bankruptcy introduce some
12
For examples of theoretical work on endogenizing the incompleteness of markets through limited enforcement of
contracts or asymmetric information, see Kehoe and Levine (1993) and Cole and Kocherlakota (2001).
13
Consider an individual owner of a business that generates a stream of risky prots. If the rms only external
nance is non-contingent debt then the individual bears all the risk (except in the case of default). If the individual
wanted to share risk with other investors then one possibility would be to replace the non-contingent debt with
state-contingent bonds where the payos on these bonds are positively related to the rms prots. However, what
is commonly observed is not issuance of state-contingent bonds but equity nancing. Issuing equity also allows for
risk sharing, but unlike state-contingent bonds does not need to spell out a schedule of payments in all states of the
world. There is no right to any specic payment in any specic state at any specic time, only the right of being
residual claimant. The lack of specic claims is balanced by control rights over the rm. However, there is no obvious
way to be residual claimant on or have control rights over a household.
21
contingency into apparently non-contingent debt contracts. However, default and bankruptcy allow
for only a crude form of contingency in extreme circumstances, and these options are not without
their costs. Renegotiation is also not costless, and evidence from consumer mortgages in both the
recent U.S. housing bust and the Great Depression suggests that the extent of renegotiation may be
ineciently low (White, 2009a, Piskorski, Seru and Vig, 2010, Ghent, 2011). Furthermore, even ex-
post ecient renegotiation of a contract with no contingencies written in ex ante need not actually
provide for ecient sharing of risk from an ex-ante perspective.
It is also possible to assess the completeness of markets indirectly through tests of the ecient
risk-sharing condition, which is equivalent to perfect correlation between consumption growth rates
of dierent households. These tests are the subject of a large literature (Cochrane, 1991, Nelson,
1994, Attanasio and Davis, 1996, Hayashi, Altonji and Kotliko, 1996), which has generally rejected
the hypothesis of full risk sharing.
Finally, even if nancial markets are incomplete, the assumption that contracts are written in
terms of specically nominal non-contingent payments is important for the analysis. The evidence
presented in Doepke and Schneider (2006) indicates that household balance sheets contain signicant
quantities of nominal liabilities and assets (for assets, it is important to account for indirect exposure
via households ownership of rms and nancial intermediaries). Furthermore, as pointed out by
Shiller (1997), indexation of private debt contracts is extremely rare. This suggests the models
assumptions are not unrealistic.
The workings of nominal GDP targeting can also be seen from its implications for ination and
the real value of nominal liabilities. Indeed, nominal GDP targeting can be equivalently described
as a policy of inducing a perfect negative correlation between the price level and real GDP, and
ensuring these variables have the same volatility. When real GDP falls, ination increases, which
reduces the real value of xed nominal liabilities in proportion to the fall in real income, and vice
versa when real GDP rises. Thus the extent to which nancial markets with non-contingent nominal
assets are suciently complete to allow for ecient risk sharing is endogenous to the monetary policy
regime: monetary policy can make the real value of xed nominal repayments contingent on the
realization of shocks. A strict policy of ination targeting would be inecient because it converts
non-contingent nominal liabilities into non-contingent real liabilities. This points to an inherent
tension between price stability and the ecient operation of nancial markets.
14
That optimal monetary policy in a non-representative-agent model should feature ination uc-
tuations is perhaps surprising given the long tradition of regarding ination-induced unpredictability
in the real values of contractual payments as one of the most important of all inations costs. As
discussed in Clarida, Gal and Gertler (1999), there is a widely held view that the diculties this
induces in long-term nancial planning ought to be regarded as the most signicant cost of ina-
tion, above the relative price distortions, menu costs, and deviations from the Friedman rule that
have been stressed in representative-agent models. The view that unanticipated ination leads to
14
In a more general setting where the incompleteness of nancial markets is endogenized, ination uctuations
induced by nominal GDP targeting may play a role in minimizing the costs of contract renegotiation or default when
the economy is hit by an aggregate shock.
22
inecient or inequitable redistributions between debtors and creditors clearly presupposes a world
of incomplete markets, otherwise ination would not have these eects. How then to reconcile this
argument with the result that the incompleteness of nancial markets suggests nominal GDP target-
ing is desirable because it supports ecient risk sharing? (again, were markets complete, monetary
policy would be irrelevant to risk sharing because all opportunities would already be exploited)
While nominal GDP targeting does imply unpredictable ination uctuations, the resulting real
transfers between debtors and creditors are not an arbitrary redistribution they are perfectly cor-
related with the relevant fundamental shock: unpredictable movements in aggregate real incomes.
Since future consumption uncertainty is aected by income risk as well as risk from uctuations in
the real value of nominal contracts, it is not necessarily the case that long-term nancial planning
is compromised by ination uctuations that have known correlations with the economys funda-
mentals. An ecient distribution of risk requires just such uctuations because the provision of
insurance is impossible without the possibility of ex-post transfers that cannot be predicted ex ante.
Unpredictable movements in ination orthogonal to the economys fundamentals (such as would
occur in the presence of monetary-policy shocks) are inecient from a risk-sharing perspective, but
there is no contradiction with nominal GDP targeting because such movements would only occur if
policy failed to stabilize nominal GDP.
It might be objected that if debtors and creditors really wanted such contingent transfers then
they would write them into the contracts they agree, and it would be wrong for the central bank to try
to second-guess their intentions. But the absence of such contingencies from observed contracts may
simply reect market incompleteness rather than what would be rationally chosen in a frictionless
world. Reconciling the non-contingent nature of nancial contracts with complete markets is not
impossible, but it would require both substantial dierences in risk tolerance across households
and a high correlation of risk tolerance with whether a household is a saver or a borrower. With
assumptions on preferences that make borrowers risk neutral or savers extremely risk averse, it
would not be ecient to share risk, even if no frictions prevented households writing contracts that
implement it.
There are a number of problems with this alternative interpretation of the observed prevalence of
non-contingent contracts. First, there is no compelling evidence to suggest that borrowers really are
risk neutral or savers are extremely risk averse relative to borrowers. Second, while there is evidence
suggesting considerable heterogeneity in individuals risk tolerance (Barsky, Juster, Kimball and
Shapiro, 1997, Cohen and Einav, 2007), most of this heterogeneity is not explained by observable
characteristics such as age and net worth (even though many characteristics such as these have
some correlation with risk tolerance). The dispersion in risk tolerance among individuals with
similar observed characteristics also suggests there should be a wide range of types of nancial
contract with dierent degrees of contingency. Risk neutral borrowers would agree non-contingent
contracts with risk-averse savers, but contingent contracts would be oered to risk-averse borrowers.
Another problem with the complete markets but dierent risk preferences interpretation relates
to the behaviour of the price level over time. While nominal GDP has never been an explicit target
of monetary policy, nominal GDP targetings implication of a countercyclical price level has been
23
largely true in the U.S. during the post-war period (Cooley and Ohanian, 1991), albeit with a
correlation coecient much smaller than one in absolute value, and a lower volatility relative to
real GDP. Whether by accident or design, U.S. monetary policy has had to a partial extent the
features of nominal GDP targeting, resulting in the real values of xed nominal payments positively
co-moving with real GDP (but by less) on average. In a world of complete markets with extreme
dierences in risk tolerance between savers and borrowers, ecient contracts would undo the real
contingency of payments brought about by the countercyclicality of the price level, for example,
through indexation clauses. But as discussed in Shiller (1997), private nominal debt contracts have
survived in this environment without any noticeable shift towards indexation. Furthermore, both
the volatility of ination and correlation of the price level with real GDP have changed signicantly
over time (the high volatility 1970s versus the Great Moderation, and the countercyclicality of
the post-war price level versus its procyclicality during the inter-war period). The basic form of
non-contingent nominal contracts has remained constant in spite of this change.
15
Finally, while the policy recommendation of this paper goes against the long tradition of citing
the avoidance of redistribution between debtors and creditors as an argument for price stability,
it is worth noting that there is a similarly ancient tradition in monetary economics (which can be
traced back at least to Bailey, 1837) of arguing that money prices should co-move inversely with
productivity to promote fairness between debtors and creditors. The idea is that if money prices
fall when productivity rises, those savers who receive xed nominal incomes are able to share in
the gains, while the rise in prices at a time of falling productivity helps to ameliorate the burden
of repayment for borrowers. This is equivalent to stabilizing the money value of incomes, in other
words, nominal GDP targeting. The intellectual history of this idea (the productivity norm) is
thoroughly surveyed in Selgin (1995). Like the older literature, this paper places distributional
questions at the heart of monetary policy analysis, but studies policy through the lens of mitigating
ineciencies in incomplete nancial markets, rather than with looser notions of fairness.
3 Incomplete nancial markets in a monetary DSGE model
3.1 Households
The economy contains a measure-one population of households. Time is discrete and households
are innitely lived. There are equal numbers of two types of households, referred to as borrowers
and savers and indexed by type {b, s}. A representative household of type has preferences
given by the following utility function
U
,t
=

=0
E
t
_
_
_
1

=1

,t+
_
_
_
C
1
,t+
1

H
1+
1

,t+
1 +
1

_
_
_
_
, [3.1]
15
It could be argued that part of the reluctance to adopt indexation is a desire to avoid eliminating the risk-sharing
oered by nominal contracts when the price level is countercyclical.
24
where C
,t
is per-household consumption of a composite good by type- households at time t, and
H
,t
is hours of labour supplied. The two types are distinguished by their subjective discount factors,
with
,t
being the discount factor of type- households between time t and t + 1. Both types have
a constant coecient of relative risk aversion given by , and a constant elasticity of intertemporal
substitution given by
1
. The household-specic Frisch elasticity of labour supply is

.
Each household of type receives real income Y
,t
at time t, to be specied below. The discount
factor
,t
of type- households is assumed to be the following:

,t
=

_
C
,t
Y
,t
_
, where

(c) =

c
(1)
, [3.2]
and where the parameters
b
,
s
, and are such that 0 <
b
<
s
< and 0 < < 1. It is
assumed individual households of type take
,t
as given, that is, they do not internalize the eect
of their own consumption on the discount factor.
There are two dierences compared to a representative-household model with a standard time-
separable utility function. First, there is heterogeneity in discount factors because borrowers
are more impatient than savers (
b
<
s
), all else equal. This is the key assumption that will
give rise to borrowing and saving in equilibrium by the households that have been referred to as
borrowers and savers. Second, discount factors display the marginal increasing impatience ( < 1)
property of Uzawa (1968), in that the discount factor is lower when consumption is higher (relative
to income), all else equal. This assumption is invoked for technical reasons because it ensures the
wealth distribution will be stationary around a well-dened non-stochastic steady state.
16
That
households take discount factors as given is assumed for simplicity and is analogous to models of
external habits (see for example, Abel, 1990).
17
The composite good C
,t
is a CES aggregate of a measure-one continuum of dierentiated goods
indexed by [0, 1]. The aggregator is the same for both types of households and features a
constant elasticity of substitution between goods. Households allocate spending C
,t
() between
goods to minimize the nominal expenditure P
t
C
,t
required to obtain C
,t
units of the consumption
aggregator:
P
t
C
,t
= min
{C
,t
()}
_
[0,1]
P
t
()C
,t
()d s.t. C
,t
=
__
[0,1]
C
,t
()
1

d
_
1
, [3.3]
where P
t
() is the nominal price of good .
Households of type face a real wage w
,t
for their labour. All households own equal (non-
tradable) shareholdings in a measure-one continuum of rms, with rm paying real dividend J
t
().
All households are assumed to face a common lump-sum tax T
t
in real terms. Real disposable
16
None of the qualitative results depends on being signicantly below one ( = 1 is the standard case of xed
discount factors), and moreover, the quantitative importance of the results does not vanish when is arbitrarily close
to one. The assumption is analogous to those employed in small open-economy models to ensure a stationary net
foreign asset position (see Schmitt-Grohe and Uribe, 2003). An alternative is to work with an overlapping generations
model where the utility function is entirely standard, which automatically has a stationary wealth distribution because
households have nite lives. This avenue was explored in an earlier working paper (Sheedy, 2013).
17
The assumption can be relaxed at the cost of more complicated algebra, but there is little impact on the results
when is close to one. The changes to the results are described in the appendix.
25
income for households of type is thus
Y
,t
= w
,t
H
,t
+
_
[0,1]
J
t
()d T
t
. [3.4]
3.2 Incomplete nancial markets
The only liability that can be issued by households is a non-contingent nominal bond. Households
can take positive or negative positions in this bond (save or borrow), and there is no limit on
borrowing other than being able to repay in all states of the world given non-negativity constraints
on consumption. With this restriction, no default will occur, and thus bonds are risk-free in nominal
terms.
18
The nominal bond has the following structure. One newly issued bond at time t makes a stream
of coupon payments in subsequent time periods, paying 1 unit of money (a normalization) at time
t + 1, then units at t + 2,
2
at t + 3, and so on (0 < ). The geometric structure of
the coupon payments means that a bond issued at time t is after its time-t coupon payment
equivalent to a quantity

of new date-t bonds. If Q

t
denotes the price in terms of money at time
t of one new bond then the absence of arbitrage opportunities requires that bonds issued at date
t have price

Q
t
at time t. It therefore suces to track the overall quantity of bonds in terms
of new-bond equivalents, rather than the quantities of each vintage separately.
19
The ow budget identity at time t of households of type is:
C
,t
+
Q
t
B
,t
P
t
= Y
,t
+
(1 +Q
t
)B
,t1
P
t
, [3.5]
where B
,t
denotes the outstanding quantity of bonds (in terms of new-bond equivalents) held (or
issued, if negative) by type- households at the end of period t. The term 1 + Q
t
refers to the
coupon payment plus the resale value of bonds acquired or issued in the past.
3.3 Firms
Firm [0, 1] is the monopoly producer of dierentiated good . Goods are produced using an
aggregator of labour inputs. Production of good is denoted by Y
t
(), and rm s labour usage by
H
t
(), and w
t
denotes the wage cost per unit of H
t
(). The rm pays out all real prots at time t as
dividends J
t
():
J
t
() =
P
t
()
P
t
Y
t
() w
t
H
t
(), where Y
t
() = A
t
H
t
()
1
1+
, and Y
t
() =
_
P
t
()
P
t
_

C
t
. [3.6]
The rst equation following the denition of prots is the production function, with A
t
denoting
the common exogenous productivity level, and where the parameter determines the extent of
diminishing returns to labour ( 0). The nal equation in [3.6] is the demand function that arises
from the household expenditure minimization problem [3.3].
18
The model abstracts from the choice of default when repayment is feasible.
19
Woodford (2001) uses this modelling device to study long-term government debt. See Garriga, Kydland and

Sustek (2013) for a richer model of mortgage contracts.

26
The labour input H
t
() is an aggregator of labour supplied by the two types of households. Firms
are assumed to receive a proportional wage-bill subsidy at rate
1
. Firms choose labour inputs
H
,t
() to minimize the post-subsidy cost w
t
H
t
() of obtaining a unit of the aggregate labour input
H
t
():
w
t
H
t
() = min
H
,t
()
(1
1
) (w
b,t
H
b,t
+w
s,t
H
s,t
) s.t. H
t
() = 2H
b,t
()
1
2
H
s,t
()
1
2
. [3.7]
The labour aggregator has a Cobb-Douglas functional form, implying a unit elasticity of substitution
between dierent labour types.
20
3.4 Sticky prices
Price adjustment is assumed to be staggered according to the Calvo (1983) pricing model. In each
time period, there is a probability that rm must continue to use its previous nominal price
P
t1
(). If at time t a rm does receive an opportunity to change price, it sets a reset price denoted
by

P
t
. The reset price is set to maximize the current and expected future stream of prots. Future
prots conditional on continuing to charge

P
t
are multiplied by the probability

that the reset

price will actually remain in use periods ahead, and are discounted using the real interest rate
t
:
max

P
t

=0
E
t
_
_
_

1
=0
(1 +
t+
)
_
_
_
_

P
t
P
t+
_
1

w
t+
C

t+
A
1+
t
_

P
t
P
t+
_
(1+)
_
_
C
t+
_
_
. [3.8]
3.5 Money and monetary policy
The economy is cash-less in that money is not required for transactions, but money is used as a
unit of account in writing nancial contracts and in pricing goods. Monetary policy is assumed to
be able to determine a path for the price level P
t
.
3.6 Fiscal policy
The only role of scal policy here is to provide the wage-bill subsidy to rms by collecting equal
amounts of a lump-sum tax from all households. It is assumed the scal budget is in balance, so
taxes T
t
are set at the level required to fund the current subsidy:
21
T
t
=
1
_
[0,1]
(w
b,t
H
b,t
() +w
s,t
H
s,t
()) d. [3.9]
20
The use of an aggregator of dierent labour types is the standard approach to studying the welfare costs of sticky
wages (Erceg, Henderson and Levin, 2000). Here, a unit elasticity of substitution is the most analytically convenient
assumption, though a priori it is not clear whether the labour of dierent household types is more substitutable or
more complementary than this. In the model, being a borrower is simply a matter of being more impatient than
savers, but empirically, the average borrower is likely to dier from the average saver in respects such as age that
might mean their labour is not perfectly substitutable.
21
The wage-bill subsidy is a standard assumption which ensures the economys steady state is not distorted (Wood-
ford, 2003). A balanced-budget rule is assumed to avoid any interactions between scal policy and nancial markets.
27
3.7 Market clearing
Market clearing in goods, labour, and bond markets requires:
1
2
C
b,t
() +
1
2
C
s,t
() = Y
t
(), for all [0, 1]; [3.10a]
_
[0,1]
H
,t
()d =
1
2
H
,t
, for all {b, s}; [3.10b]
1
2
B
b,t
+
1
2
B
s,t
= 0. [3.10c]
3.8 Equilibrium
The derivation of the equilibrium conditions is presented in the appendix. The analysis of incomplete
nancial markets follows the method used for the simple model in section 2, while other aspects of
the model are standard features of New Keynesian models with sticky prices.
The debt-to-GDP ratio d
t
, the loans-to-GDP ratio l
t
, and the ex-post real return r
t
are dened
as follows:
D
t

1
2
(1 +Q
t
)B
b,t1
P
t
, L
t

1
2
Q
t
B
b,t
P
t
, and 1 +r
t

(1 +Q
t
)P
t1
Q
t1
P
t
, [3.11]
and the ratios of consumption to GDP are c
,t
C
,t
/Y
t
. The yield-to-maturity j
t
is dened by:
Q
t
=

=1

1
(1 +j
t
)

, implying j
t
=
1
Q
t
1 +. [3.12]
28
With these denitions in hand, the full set of equilibrium conditions is collected below:

t
= E
t
r
t+1
; [3.13a]
d
t
=
_
1 +r
t
1 +g
t
_
l
t1
; [3.13b]
c
b,t
= 1 2(d
t
l
t
), and c
s,t
= 1 + 2(d
t
l
t
); [3.13c]
1 =
,t
E
t
_
(1 +r
t+1
)(1 +g
t+1
)

_
c
,t+1
c
,t
_

_
; [3.13d]

,t
=

c
(1)
,t
; [3.13e]
lim

_
1

=0

,t+
(1 +g
t+1+
)
1
_
c

,t+
l
t+
= 0; [3.13f]
1 +r
t
=
_
1 +j
t
1 +
t
__
1 +j
t1
1 +j
t
_
; [3.13g]
lim

E
t
__

=1

,t+1
(1 +g
t+
)

(1 +
t+
)
_
c
,t+
c
,t+1
_

_
(1 +j
t+
)
1
_
= 0; [3.13h]
1 +g
t
=
Y
t
Y
t1
; [3.13i]
x
t
= (1
1
)(1 +)

1+

t
Y
++
1+

t
A
1++
1+

t
_
c

b
1+
b
b,t
c

s
1+
s
s,t
_

b
1+
b
+

s
1+
s
; [3.13j]

t
=
_
_
(1 +
t
)
(1+)

1+
t1
+ (1 )
_
1 (1 +
t
)
1
1
_
(1+)
1
_
_
1
1+
; [3.13k]

=0
E
t
_

_
_

=1
(1 +g
t+
)

1
=0
(1 +
t+
)
_
_
_
_
_
1(1+
t
)
1
1
_1+
1

=1
(1 +
t+
)
1

1
x
t+

=1
(1 +
t+
)
(1+)
_
_
_
_

_
= 0. [3.13l]
The variable x
t
is the level of real marginal cost for a rm whose good sells at a price equal to the
general price level, and
t
is the eect of relative-price distortions on aggregate productivity.
In a steady state where exogenous productivity A
t
is growing at a constant rate, there is a
steady-state rate of real GDP growth g. The steady-state consumption-income ratios are given by
c
b
= 1 , c
s
= 1 +, where
1 (
b
/
s
)
(1)
1 + (
b
/
s
)
(1)
, [3.14]
where 0 < < 1. The term depends on the relative patience
b
/
s
of the two household types
and the utility-function parameters and . The steady-state discount factors and real interest
rate are:

b
=

s
=
__

1
(1)
b
+
1
(1)
s
_
_
2
_
(1)
, = r =
1 + g

1, where (1 + g)
1
,
[3.15]
and it is assumed g is low enough to ensure that 0 < < 1. In the steady state, the discount factors
29
of the two types are aligned at , which is eectively an average of the patience parameters
b
and

s
. The steady-state debt-to-GDP ratio can be written in terms of and as follows:

d =

2(1 )
, and

l =

2(1 )
. [3.16]
The model can be parameterized directly with and rather than the two patience parameters
b
and
s
(leaving , , and g to be chosen separately). The term plays the usual role of the discount
factor in a representative-household economy given its relationship with the real interest rate (with
an adjustment for steady-state real GDP growth). The term quanties the extent of heterogeneity
between borrower and saver households, which is related to the amount of borrowing and saving
that occurs in equilibrium, and hence to the debt-to-GDP ratio in [3.16]. Given equation [3.14],
can be interpreted as the debt service ratio because it is the net fraction of income transferred by
borrowers to savers.
22
As will be seen, is a sucient statistic for the extent of heterogeneity in
the economy, with 0 being the limiting case of a representative-household economy (
b

s
).
Finally, rather than specify the bond coupon parameter directly, the steady-state fraction of debt
that is not renanced each period is taken to be a parameter. This fraction is = /((1+ )(1+ g)).
4 Optimal monetary policy
This section studies the features of optimal monetary policy in the full model of section 3. The rst
step is to consider what the equilibrium of the economy would be in the absence of two key frictions:
incomplete nancial markets and sticky prices.
4.1 A rst-best benchmark
Consider an economy with complete nancial markets, but which is otherwise identical to that
described in section 3 (in particular, this economy may have sticky prices). The equilibrium
consumption-GDP ratios and discount factors in this hypothetical economy are
c

b,t
= 1 , c

s,t
= 1 +, and

b,t
=

s,t
= , [4.1]
which are identical to the steady-state values of these variables from [3.14] and [3.15]. As expected,
this equilibrium features full risk sharing between borrowers and savers. The associated level of the
debt-to-GDP ratio is
d

t
=

2
E
t
_

=0

=1
(1 +g
t+
)
1
_
. [4.2]
The complete-markets debt-to-GDP ratio d

t
is referred to as the natural debt-to-GDP ratio. This
terminology is motivated by analogy with such concepts as the natural rate of interest, the natural
rate of unemployment, and the natural level of output. In common with these other concepts, it
represents an economic outcome that would be achieved in the absence of a particular friction (here,
22
Strictly speaking, when real GDP growth is dierent from zero, is the debt service ratio net of new borrowing.
30
the friction is the inability to issue state-contingent bonds, in contrast to the frictions of imperfect
information or nominal rigidities that aect the supply of output or labour in many models).
A generalization of this concept is the equilibrium of an economy where complete nancial
markets are open only from some date t
0
onwards, taking as given the initial wealth distribution at
t
0
1. The consumption ratios in this case are denoted by c

,t|t
0
, and the associated debt-to-GDP
ratio by d

t|t
0
. The values of c

,t|t
0
depend only on the predetermined wealth distribution at t
0
1.
Now consider an economy with both complete nancial markets (open from t
0
onwards) and
fully exible prices (the case where = 0). The equilibrium of this economy represents one rst-
best allocation starting from date t
0
. Given the complete-markets consumption ratios c

,t|t
0
(which
depend only on the wealth distribution at t
0
1), the equilibrium level of output

Y

t|t
0
can be obtained,
which is referred to as the natural level of output:

t|t
0
=
_
1
1 +
_
c

b
1+
b
b,t|t
0
c

s
1+
s
s,t|t
0
_

b
1+
b
+

s
1+
s
A
1++
1+

t
_
1
++
1+

. [4.3]
It is instructive to consider the special cases of fully exible prices or complete nancial markets.
First suppose nancial markets are incomplete but prices are fully exible. Monetary policy can
be used to replicate complete nancial markets as in the model of section 2, which implies that
households make the same labour supply decisions as they would in a complete-markets economy.
Since prices are fully exible, this does not lead to any relative price distortions, and the aggregate
level of output is ecient.
The second special case is that of complete nancial markets or a representative household
( = 0), but where prices are sticky. Because nancial markets are complete, full risk sharing occurs
without any policy intervention. In this case, strict ination targeting (with a zero ination target,

t
= 0) avoids any relative-price distortions and ensures the level of aggregate output is what would
prevail were prices were fully exible, which is ecient.
Thus, with exible prices, there is no trade-o between a policy that supports risk sharing and
achieving the optimal level of aggregate output. With complete nancial markets, there is no trade-
o between avoiding relative-price distortions and achieving the optimal level of aggregate output.
However, with both incomplete nancial markets and sticky prices, all three objectives of risk
sharing, avoiding relative-price distortions, and closing the output gap are in conict. These trade-
os are studied by deriving a loss function to approximate the welfare function, and approximations
of the equilibrium conditions that represent the constraints on monetary policy.
4.2 Policy tradeos
An exact analytical solution for optimal monetary policy is not available, so this section resorts to
nding the log-linear approximation (the rst-order perturbation around the non-stochastic steady
state), which can be found analytically. The notational convention is that variables in a sans
serif font denote log deviations of the equivalent variables in roman letters from their steady-state
values (log deviations of interest rates, ination rates, and growth rates are log deviations of the
31
corresponding gross rates; for variables that have no steady state, the sans serif letter simply denotes
the logarithm of that variable). It is also convenient to dene the debt gap

d
t|t
0
d
t
/d

t|t
0
, the
deviation of the actual debt-to-GDP ratio from its value with complete nancial markets, and the
output gap

Y
t|t
0
Y
t
/

t|t
0
.
A parameter restriction on the Frisch elasticities of borrowers and savers is imposed that implies
the wealth distribution has no eect on aggregate labour supply (up to a rst-order terms):

b
= (1 )/(1 +), and
s
= (1 +)/(1 ), [4.4]
where is the steady-state debt service ratio dened in [3.14] and is the eective aggregate Frisch
elasticity of labour supply (it is assumed that < 1/). In practice, this assumption does not
make much dierence to the results, but it does simplify the analysis and make the model easier to
compare to a standard New Keynesian model. One implication is that the growth rate of output
with exible prices g

t
does not depend on the initial wealth distribution, so the t
0
subscript can
be dropped here, and also from the output gap

Y
t
. The case where the distribution of wealth does
aect aggregate labour supply is considered in the appendix.
Using the Pareto weights

|t
0
= c

,t
0
|t
0
/

Y

1
t
0
that support the rst-best equilibrium with
exible prices and complete nancial markets (from t
0
onwards), a second-order approximation
(around a steady state with zero ination) of the welfare function is given below in terms of the loss
function L
t
0
:
L
t
0
=
1
2

t=t
0

tt
0
E
t
0
_

2
(1 )
2
(1
2
)(1 )
2
_
1 +

(1 +)(1 +)
_

d
2
t|t
0
+
(1 +)
(1 )(1 )

2
t
+
_
+ +
1 +

Y
2
t
_
. [4.5]
The units of the loss function are percentage equivalents of initial output. The loss function in-
cludes the squared debt-to-GDP gap

d
t|t
0
, which is a sucient statistic for the welfare loss due to
deviations from complete markets. The coecient is increasing in risk aversion and the degree
of heterogeneity between borrower and saver households as measured by . Intuitively, greater risk
aversion increases the importance of the risk sharing found in complete nancial markets, while
greater heterogeneity between households leads to larger nancial positions of borrowers and savers,
so a given percentage change in nancial wealth has a larger impact on consumption.
The second term in the loss function [4.5] is the squared ination rate, which is a sucient
statistic for the welfare loss due to relative-price distortions. This is a well-known property of Calvo
pricing, and the coecient in the loss function is the same as in standard models (see Woodford,
2003). The coecient of ination is increasing in the price elasticity of demand because a higher
price sensitivity implies that a given amount of price dispersion now leads to greater dispersion of
the quantities produced of dierent goods for which preferences and production technologies are
identical. The coecient is increasing in price stickiness because longer price spells imply that a
given amount of ination leads to greater relative-price distortions. The coecient is also increasing
in the parameter (the output elasticity of marginal cost), which will be seen to determine the
32
degree of real rigidity in the economy. Both nominal and real rigidity increase the welfare costs of
ination.
The third term in the loss function is the squared output gap. This represents the losses from
output and employment deviating from their ecient levels. The coecient is also the same as
found in a standard New Keynesian model.
Since the steady state of the economy is not distorted (there are no linear terms in the loss
function [4.5]), a rst-order accurate approximation of optimal monetary policy can be obtained
by minimizing the loss function subject to rst-order accurate approximations of the economys
equilibrium conditions. The relevant constraints are given below:
(
t
E
t

t+1
) =

Y
t
, where
(1 +)
(1 )(1 )
, and + +
1 +

; [4.6a]
E
t

d
t+1|t
0
=

d
t|t
0
; [4.6b]
j
t1
j
t
1

d
t|t
0
+

d
t1|t
0

Y
t|t
0
+

Y
t1|t
0

t

(1 )(1 )

(
t
E
t1

t
) = r

t
;
[4.6c]
lim

()

E
t
j
t+
= 0, [4.6d]
where r

t
is a variable that depends only on the exogenous growth rate g

t
of the natural level of
output:
r

t
= g

t
+ (1 )

=0

(E
t
g

t+
E
t1
g

t+
). [4.7]
The rst constraint [4.6a] is the standard New Keynesian Phillips curve linking ination, expec-
tations of future ination, and the output gap. The coecients are the same as in the standard
New Keynesian model. To understand the constraints [4.6b][4.6d] specic to the incompleteness of
nancial markets, rst consider the special case where the Frisch elasticity of labour supply is zero,
in which case the output gap

Y
t
is always zero.
4.3 Incomplete nancial markets versus relative-price distortions
With a zero Frisch elasticity of labour supply ( = 0), the Phillips curve [4.6a] reduces to

Y
t
= 0
(since ). In this case, the constraint in [4.6c] simplies to:
j
t1
j
t
1

d
t|t
0
+

d
t1|t
0

t
= r

t
. [4.8]
The constraint [4.6b] determines the predictable component of the debt-to-GDP gap

d
t|t
0
, and [4.8]
links ination
t
and nominal bond yields j
t
to the unexpected component of the debt-to-GDP gap.
The nal constraint [4.6d] is simply a transversality condition on the bond price or yield. Optimal
monetary policy minimizes the loss function [4.5] (ignoring here the term in the output gap

Y
t
)
subject to [4.6b], [4.8], and [4.6d].
33
The optimal monetary policy features uctuations in the debt-to-GDP gap:

d
t|t
0
=

d
t1|t
0
(1 )
t
, with
t
=

=0

_
1 (1

)
_
(E
t
g
t+
E
t1
g
t+
) , [4.9a]
and =
_
1 +
(1 +)(1
2
)(1 )
2
(1
2
)(1
2
)

2
(1 )(1 )(1 )
2
_
1
. [4.9b]
Optimal policy also features ination uctuations, with optimal ination persistence determined by
the maturity of debt contracts (the parameter ):

t
=
_
_
_

t1
(1
2
)
t
if t t
0
+ 1
(1
2
)
t
if t = t
0
. [4.10]
Fluctuations in the growth rate g

t
of exible-price output (due to shocks to TFP) lead to changes
in the ex-post real return on the complete-markets portfolio. To replicate complete nancial markets,
the central bank needs to vary ination and nominal bond yields so as to mimic this real return.
Overall, what matters are unexpected changes in the discounted sum of current and future growth
rates, adjusted for any mitigating (or aggravating) changes in real interest rates. This is the shock

t
given in [4.9a]. The term (1

) is the adjustment for changes in real interest rates caused by

revised expectations of the economys future growth prospects. The parameter is the elasticity
of the real interest rate with respect to expected real GDP growth. Since changes in (ex-ante) real
interest rates only matter to the extent that debt is renanced, for growth expectations periods
ahead, the interest-rate eect is proportional to the fraction 1

of existing debt that will be

renanced by then.
With sticky prices, replicating complete nancial markets through variation in ination is now
costly, so the central bank tolerates some deviation from complete markets.
23
To the extent that in
[4.9b] is less than one, a shock
t
leads to uctuations in the debt-to-GDP gap

d
t
. These uctuations
are persistent because of the constraint [4.6b]: the serial correlation of the debt-to-GDP gap is .
Once a non-zero debt gap arises at time t, there is no predictable future policy action that can undo
its future consequences.
If were equal to 1, equation [4.9a] shows that the debt-to-GDP gap would be completely
stabilized, and if were 0, equation [4.10] shows that ination would be completely stabilized.
Since 0 < < 1, optimal monetary policy can be interpreted as a convex combination of strict
ination targeting and a policy that replicates complete nancial markets. As the responses of

d
t|t
0
and
t
to the shock
t
are linearly related to , the terms and 1 can be interpreted respectively
as the weights on completing nancial markets and avoiding relative-price distortions. Comparing
equations [4.5] and [4.9b], it can be seen that is positively related to the ratio of the coecients
of

d
2
t|t
0
and
2
t
in the loss function divided by (1
2
) and (1
2
).
Greater risk aversion () or more heterogeneity and hence more borrowing () increase the
23
If the nominal rigidity were sticky wages rather than sticky prices, the central bank would care about nominal
wage ination rather than price ination (Erceg, Henderson and Levin, 2000). In that case, the tension with the goal
of replicating complete nancial markets is reduced because targeting nominal income growth is less likely to be in
conict with stabilizing nominal wage ination than stabilizing price ination.
34
coecient of

d
2
t|t
0
and thus ; a larger price elasticity of demand (), stickier prices (), or more real
rigidities () increase the coecient of
2
t
and thus reduce . The optimal trade-o is also aected
by the constraints in [4.6b] and [4.8], which explain the presence of the terms (1
2
) and (1
2
)
in the formula for . A greater value of increases the persistence of the debt-to-GDP gap, which
makes uctuations in

d
t|t
0
more costly than suggested by the loss function coecient alone. The
parameter aects the link between bond yields and the debt-to-GDP gap. It is seen that an
increase in leads to a higher value of , the intuition for which is related to the optimal behaviour
of ination. Finally, note that while the optimal policy responses depend on the stochastic process
for real GDP growth, the optimal weight does not.
Equation [4.10] shows that optimal monetary policy features ination persistence with serial
correlation given by the debt maturity parameter = /(1 + g). The steady-state fractions of
existing and newly issued debt are and 1 respectively, so the result is that ination should
return to its average value at the same rate at which debt is renanced. At the extremes, one-period
debt ( = 0 and = 0) corresponds to serially uncorrelated ination, while perpetuities ( = 1, for
which 1) correspond to near random-walk persistence of ination.
To understand this, note that with one-period debt, the current bond yield j
t
disappears from
the constraint [4.8], thus the only way that policy can aect

d
t|t
0
is through an unexpected change
in current ination. With longer maturity debt, the range of policy options increases. Changes
in current bond yields j
t
are also relevant in addition to current ination, and the bond yield is
aected by expectations of future ination. The three constraints [4.6b], [4.8], and [4.6d] imply

d
t|t
0
=

d
t1|t
0

=0
()

(E
t

t+
E
t1

t+
), where

indicates the fraction of existing debt

that will not have been renanced after time periods. This shows it is now possible to use ination
that is spread out over time to inuence the debt-to-GDP gap

d
t|t
0
, not only ination surprises.
Furthermore, this ination smoothing is optimal because the welfare costs of ination are
convex (ination appears in the loss function [4.5] as
2
t
), so the costs of a given cumulated amount
of ination are smaller when spread out over a number of quarters or years than when all the
ination occurs in just one quarter. This is analogous to the tax smoothing argument of Barro
(1979). Interestingly, the argument shows that high degrees of ination persistence need not be
interpreted as a failure of policy. Dierently from the tax smoothing argument, it is generally not
optimal for ination to display random walk or near-random walk persistence unless debt contracts
are close to perpetuities. As the maturity parameter is reduced and thus falls signicantly
below one, expectations of ination far in the future have a smaller eect on bond yields than
ination in the near future. The further in the future ination is expected to occur, the less eective
it is at inuencing real returns and thus the debt-to-GDP ratio.
Even if optimal monetary policy places a substantial weight on the problem of incomplete
markets compared to relative-price distortions, in what sense does monetary policy still resemble
a nominal GDP target? Optimal ination smoothing E
t

t+1
=
t
(from [4.10]) pins down the
predictable component of the ination rate, but this in itself is not a complete description of pol-
icy because it leaves unspecied the unexpected component of ination. Together with ination
smoothing, specifying how much ination reacts to a shock on impact, or equivalently, how much
35
cumulated ination will follow a shock, completely characterizes the path of all nominal variables.
It turns out that optimal monetary policy retains the essential feature of nominal GDP targeting
in generating a negative comovement between prices and output. However, because of the desire to
smooth ination, the central bank should not aim to stabilize nominal GDP (or a weighted measure
of nominal GDP) exactly on a quarter-by-quarter basis. Instead, optimal policy can be formulated
as a long-run target for weighted nominal GDP. When real GDP is non-stationary because TFP
follows a non-stationary process, optimal monetary policy features cointegration between the price
level and output. This cointegrating relationship can be interpreted a long-run target for weighted
nominal GDP because there is some linear combination of prices and real output that is eventually
returned to a constant or a deterministic trend following a shock to output.
4.4 The ex-ante real interest rate and the output gap
In the general case where labour supply is elastic ( > 0), monetary policy can also aect the
output gap and the ex-ante real interest rate. This is important not only because of a concern for
stabilizing the output gap, but also because the output gap aects the debt-to-GDP gap through
the constraint [4.6c]. The main reason for this is the connection between the output gap and the
ex-ante real interest rate
t
. The real interest rate is given by
t
= E
t

Y
t+1

Y
t
+

t|t
0
, where

t
= E
t
[

t+1

t
] is the real interest rate at the natural level of output. This is the source of the
coecient of

Y
t|t
0
in [4.6c].
The problem of minimizing the loss function [4.5] subject to the constraints [4.6a][4.6d] (with
commitment at some distant past date t
0
) has a solution with the following properties:
the debt gap

d
t
is an AR(1) process with an innovation proportional to
t
=

=0

(1 (1

))(E
t
g

t+
E
t1
g

t+
) and an autoregressive root ; ination
t
is an ARMA(2,1) process with an
innovation proportional to
t
and autoregressive roots equal to and , where the latter is:
=
2
1 + +

+
_
_
1 + +

_
2
4
, [4.11]
which satises 0 < < 1. With strict ination targeting, the debt gap

d
t
is an AR(1) process with
an innovation proportional to
t
and an autoregressive root . The solution for the debt gap is
therefore a multiple of the debt gap under the optimal policy, and the optimal policy weight on
incomplete nancial markets is dened so that the debt gap under the optimal policy is equal to
the debt gap under strict ination targeting multiplied by 1 .
To understand the new aspects of the optimal monetary policy problem with elastic labour
supply, note that policy can now inuence three variables which have implications for the debt gap
and thus the extent to which complete nancial markets are replicated: ination, real GDP, and the
ex-ante real interest rate. Ination aects the ex-post real return on nominal bonds and thus the
value of existing debt (as before). Real GDP (and hence the output gap) aects the denominator
of the debt-to-GDP ratio. The ex-ante real interest rate aects the ongoing costs of servicing debt
relative to the stream of current and future labour income (formally, the ex-ante real interest rate
36
inuences the debt gap by changing the level of the debt-to-GDP ratio consistent with risk sharing).
By combining the four constraints [4.6a][4.6d], the evolution of the debt gap depends on the
shock
t
and three terms related respectively to ination, the output gap, and the ex-ante real
interest rate.
24
The eect of ination on debt is related to the unexpected change in the term
E
t
[
t
+ ()
t+1
+ ()
2

t+2
+ ] as before, where

represents the stock of debt issued in the

past that has not been renanced periods after the shock.
An increase in the output gap

Y
t
has the eect of directly boosting real GDP growth at time
t and thus reducing the debt-to-GDP ratio, but the impact on the debt gap is more subtle. Since
monetary policy has only a temporary inuence on real GDP, extra growth now reduces overall
growth in the future by exactly the same amount. Given the link between growth expectations
and the debt-to-GDP ratio consistent with risk sharing (the log linearization of equation [4.2] is
d

t
= (1 )

=1

E
t
g
t+
), the eect of the output gap on the debt gap actually depends on

Y
t
+E
t
[(

Y
t+1

Y
t
) +
2
(

Y
t+2

Y
t+1
) + ], not just

Y
t
. With the New Keynesian Phillips curve
[4.6a], it is seen that this term is equal to (1)(/)
t
, the reciprocal of the long-run Phillips curve
slope multiplied by current ination. Since it is reasonable to set the discount factor close to one
(in which case the long-run Phillips curve is close to vertical), this term is negligible for all practical
purposes (it is not exactly zero because future growth is discounted relative to current growth, so
by bringing growth forwards, there is still a small positive eect). Monetary policy therefore cannot
have a sustainable impact on the burden of debt simply through its temporary eect on the level of
real GDP.
It does not follow however that expansionary or contractionary monetary policy has no eect on
the debt burden beyond its implications for ex-post real returns through ination. There remains
the option of changing ex-ante real interest rates. Intuitively, expansionary monetary policy that
reduces the ex-ante real interest rate is eectively a transfer from savers to borrowers, what might
be labelled nancial repression. While monetary policy cannot permanently aect real interest
rates, there is no reason why cutting the real interest rates now means real interest rates in the
future must be higher than they would otherwise have been (unlike real GDP growth, as discussed
above).
Changing ex-ante real interest rates thus provides monetary policy with an alternative to in-
uencing the debt gap through the eect of ination on ex-post real returns. In contrast to the
latter, which is eective only while debt contracts are not renanced, the former is eective only
when renancing does take place. For debt renanced periods after a shock at time t, the impact
of monetary policy on the date-t debt burden is determined by the discounted sum of real interest
24
The precise expression for the evolution of the debt-to-GDP gap

d
t
is:

d
t
=

d
t1

_

t
+ (E
t
E
t1
)

=0
()

t+
+ (1 )

(E
t
E
t1
)
t
+ (1 )

(E
t
E
t1
)

=0
()

(
t+

t+1+
)
_
.
37
gaps E
t
[
+1

t+
+
+2

t++1
+
+3

t++2
+ ] from t + onwards. Given the New Keynesian
Phillips curve [4.6a] and the equation
t
= E
t
[

Y
t+1


Y
t
] linking the real interest rate and output
gaps, these terms reduce to (/)
+1
E
t
[
t++1

t+
], meaning that the slope of the ination
path over time is an indicator of nancial repression through ex-ante real interest rates. With a
(steady-state) fraction (1 )

of debt issued prior to date t being renanced at t +, the overall

eect of changes in ex-ante real interest rates on the debt burden is given by the unexpected change
in (/)(1)

=0
()

E
t
[
t++1

t+
]. Thus, the more the ination trajectory is smoothed,
the smaller is the eect of monetary policy on the ex-ante real interest rate.
As the maturity of debt increases, successful nancial repression requires an ination trajectory
with a non-zero slope further in the future. Given the Phillips curve, the required ination path
entails output gap uctuations over a longer horizon, increasing the losses from following such
a policy. Financial repression is therefore not well suited to stabilizing the debt gap when debt
contracts have a long maturity, in which case a policy of inuencing ex-post real returns through
ination smoothly spread out over time is eective at a much lower cost in terms of the implied
ination and output gap uctuations. But for short-maturity debt where only immediate ination
surprises can aect ex-post real returns, nancial repression provides an additional tool for stabilizing
the debt gap, with the losses from following this policy being small when the short-run Phillips curve
is relatively at.
5 Quantitative analysis of optimal monetary policy
This section presents a quantitative analysis of the nature of optimal monetary policy taking into
account all the features of the full model from section 3.
5.1 Calibration
Let T denote the length in years of one discrete time period in the model. The numerical results
presented here assume a quarterly frequency (T = 1/4), but the choice of frequency does not
signicantly aect the results. The parameters of the model are , , , , , , , , and . As
far as possible, these parameters are set to match features of U.S. data.
25
The baseline calibration
targets and the implied parameter values are given in Table 1 and justied below.
Consider rst the parameters and (the choice of these parameters is equivalent to specifying
the patience parameters
b
and
s
). These are calibrated to match evidence on the average price
and quantity of household debt. The price of debt is the average annual continuously compounded
real interest rate r paid by households for loans. As seen in [3.15], the steady-state growth-adjusted
real interest rate is related to . Let g denote the average annual real growth rate of the economy.
Given the length of the discrete time period in the model, 1+ = e
r T
and 1+ g = e
gT
. Hence, using
25
The source for all data referred to below is Federal Reserve Economic Data (http://research.stlouisfed.
org/fred2).
38
Table 1: Baseline calibration
Calibration targets Implied parameter values
Real GDP growth g 1.7%
Real interest rate r 5% Discount factor 0.992
Debt-to-GDP ratio D 130% Debt service ratio 8.6%
Coecient of relative risk aversion 5
Marginal propensity to consume m 6% Discount factor elasticity 0.993
Frisch elasticity of labour supply 2
Average duration of debt T
m
5 Debt maturity parameter 0.967
Price elasticity of demand 3
Marginal cost elasticity w.r.t. output 0.5
Average duration of price stickiness T
p
8/12 Calvo pricing parameter 0.625
Notes: The parameters are derived from the calibration targets using equations [5.1][5.5]. The calibration
targets are specied in annual time units; the parameter values assume a quarterly model (T = 1/4).
Sources: See discussion in section 5.1.
[3.15], can be set to:
= e
(r g)T
. [5.1]
From 1972 through to 2011 there was an average annual nominal interest rate of 8.8% on 30-year
mortgages, 10% on 4-year auto loans, and 13.7% on two-year personal loans, while the average annual
change in the personal consumption expenditure (PCE) price index over the same time period was
3.8%. The average credit-card interest rate between 1995 and 2011 was 14%. For comparison, 30-
year Treasury bonds had an average yield of 7.7% over the periods 19772001 and 20062011. The
implied real interest rates are 4.2% on Treasury bonds, 5% on mortgages, 6.2% on auto loans, 9.9%
on personal loans, and 12% on credit cards.
26
The baseline real interest rate is set to the 5% rate on
mortgages as these constitute the bulk of household debt.
27
The sensitivity analysis considers values
of r from 4% up to 7%. Over the period 19722011 used to calibrate the interest rate, the average
annual growth rate of real GDP per capita was 1.7%. Together with the baseline real interest rate
of 5%, this implies that 0.992 using [5.1]. Since many models used for monetary policy analysis
are typically calibrated assuming zero real trend growth, for comparison the sensitivity analysis also
considers values of g between 0% and 2%.
The relevant quantity variable for debt is the ratio of gross household debt to annual household
income, denoted by D.
28
This corresponds to what is dened as the loans-to-GDP ratio

l in the
26
Average PCE ination over the periods 19772001 and 20062011 was 3.5%, and 2% over the period 19952011.
27
Mortgage debt was around 77% of all household debt on average during 20062010. The baseline real interest
rate is close to the conventional calibration used in many real business cycle models (King and Rebelo, 1999). The
mortgage rate implies a spread of 0.8% between the interest rates on loans to households and Treasury bonds of the
same maturity. C urdia and Woodford (2009) consider a somewhat larger spread of 2% between interest rates for
borrowers and savers.
28
Given the heterogeneity between borrower and saver households, it would not make sense to net the nancial
assets of savers against the liabilities of borrowers. However, it might be thought appropriate to net assets and
liabilities at the level of individual households, especially since a large fraction of household borrowing is to buy
39
model (the empirical debt ratio being based on the amount borrowed rather than the subsequent
value of loans at maturity) after adjusting for the length of one time period (T years), hence D =

lT.
Using the expression for

l in equation [3.16] and given a value of , the equation can be solved for
the implied value of the debt service ratio :
=
2(1 )D
T
. [5.2]
Note that in the model, all GDP is consumed, so for consistency between the data and the models
prediction for the debt-to-GDP ratio, either the numerator of the ratio should be total gross debt
(not only household debt), or the denominator should be disposable personal income or private
consumption. Since the model is designed to represent household borrowing, and because the
implications of corporate and government debt may be dierent, the latter approach is taken.
In the U.S., like a number of other countries, the ratio of household debt to income has grown
signicantly in recent decades. To focus on the implications of levels of debt recently experienced,
the model is calibrated to match average debt ratios during the ve years from 2006 to 2010. The
sensitivity analysis considers a wide range of possible debt ratios from 0% up to 200%. Over the
period 20062010, the average ratio of gross household debt to disposable personal income was
approximately 124%, while the ratio of debt to private consumption was approximately 135%.
Taking the average of these numbers, the target chosen is a model-consistent debt-to-income ratio
of 130%, which implies using [5.2] a debt service ratio of 8.6%.
29
In estimating the coecient of relative risk aversion , one possibility would be to choose values
consistent with household portfolios of risky and safe assets. But since Mehra and Prescott (1985)
it has been known that matching the equity risk premium may require a risk aversion coecient
above 30, while values in excess of 10 are considered by many to be highly implausible.
30
Subsequent
assets (houses). If the assets held by households had the same non-contingent return as their debt liabilities then this
netting would be valid, but that is highly unlikely for assets such as housing, which experience signicant uctuations
in value. In the model, non-contingent debt is repaid only from future income, not from the sale of assets, so
the assumptions used in the calibration would be approximately correct if the value of the assets actually held by
households were positively correlated with the value of GDP and had a similar volatility. If asset returns were more
procyclical, the calibration would understate the problem of household leverage; if returns were less procyclical or
countercyclical, the calibration would overstate leverage.
29
Empirically, a direct measure of the ratio of household debt payments to disposable personal income is available,
though this is not directly comparable to the debt service ratio in the model. Between 2006 and 2010, the average
debt service ratio was 12.7%. This measure includes both interest and amortization. For conventional T
f
-year xed-
rate mortgages (where the borrower makes a sequence of equal repayments over the life of the loan) the share of
amortization in total repayments (over the life of the loan, or over all cohorts of borrowers at a point in time) is
approximately (1 e
r T
f
)/(r T
f
), where r is the annual real interest rate. Taking r 5% and T
f
= 30 years, this
formula implies that interest payments are approximately 48% of total debt service costs, yielding an estimate of the
interest-only debt-service ratio of around 6.1%. In the model, the debt service ratio is calculated only for borrowers,
not for all households as in the data, and is net of new borrowing (which is positive when there is positive income
growth). Using [3.15], the model implies the interest-only debt service ratio over all households (including the 50% of
savers with a zero debt service ratio) is given by

d/(1 + ), which is comparable to the interest-only adjustment of

the empirical measure. The baseline calibration yields a debt service ratio of approximately 6.5%, close to the 6.1%
in the data.
30
Large values of are also needed to generate a signicant ination risk premium. For illustration, suppose real
GDP follows a random walk, and the standard deviation of annual real GDP growth is set to 2.1% as found in the
data for the period 19722011. Under the exible-price optimal policy of nominal GDP targeting, the ination risk
premium would be approximately 0.044% (at an annual rate). For < 10, the ination risk premium can be no
40
analysis of the equity risk premium puzzle has attempted to build models consistent with the large
risk premium but with much more modest degrees of risk aversion.
31
Alternative approaches to estimating risk aversion have made use of laboratory experiments,
observed behaviour on game shows, and in a recent study, the choice of deductible for car insurance
policies (Cohen and Einav, 2007).
32
The survey evidence presented by Barsky, Juster, Kimball and
Shapiro (1997) potentially provides a way to measure risk aversion over stakes that are large as a
fraction of lifetime income and wealth.
33
The results suggest considerable risk aversion, but most
likely not in the high double-digit range for the majority of individuals. Overall, the weight of
evidence from studies suggests a coecient of relative risk aversion above one, but not signicantly
more than 10. A conservative baseline value of 5 is adopted, and the sensitivity analysis considers
values from zero up to 10.
34
One approach to calibrating the discount factor elasticity parameter (from [3.2]) is to select a
value on the basis of its implications for the marginal propensity to consume from nancial wealth.
Let m denote the increase in per-household (annual) consumption of savers from a marginal increase
in their nancial wealth.
35
It can be shown that m, , and are related as follows:
=
1 mT

. [5.3]
Parker (1999) presents evidence to suggest that the marginal propensity to consume from wealth lies
between 4% and 5% (for a survey of the literature on wealth and consumption, see Poterba, 2000).
However, it is argued by Juster, Lupton, Smith and Staord (2006) that the marginal propensity to
consume varies between dierent forms of wealth. They nd the marginal propensity to consume
more than 0.44%, even though in this example ination would be perfectly negatively correlated with real GDP
growth and have the same standard deviation (of 2.1%, not too far below the actual standard deviation of PCE
ination of approximately 2.5% between 1972 and 2011).
31
For example, Bansal and Yaron (2004) assume a relative risk aversion coecient of 10, while Barro (2006) chooses
a more conservative value of 4.
32
Converting the estimates of absolute risk aversion into coecients of relative risk aversion (using average annual
after-tax income as a proxy for the relevant level of wealth) leads to a mean of 82 and a median of 0.4. The stakes
are relatively small and many individuals are not far from being risk neutral, though a minority are extremely risk
averse. As discussed in Cohen and Einav (2007), the estimated level of mean risk aversion is above that found in
other studies, which are generally consistent with single-digit coecients of relative risk aversion.
33
Respondents to the U.S. Health and Retirement Study survey are asked a series of questions about whether they
would be willing to leave a job bringing in a secure income for another job with a chance of either a 50% increase
in income or a 50% fall. By asking a series of questions that vary the probabilities of these outcomes, the answers
can in principle be used to elicit risk preferences. One nding is that approximately 65% of individuals answers fall
in a category for which the theoretically consistent coecient of relative risk aversion is at least 3.8. The arithmetic
mean coecient is approximately 12, while the harmonic mean is 4.
34
The parameter is also related to the elasticity of intertemporal substitution
1
. Early estimates of intertem-
poral substitution suggested an elasticity somewhere between 1 and 2, such as those from the instrumental variables
method applied by Hansen and Singleton (1982). Those estimates have been criticized for bias due to time aggre-
gation by Hall (1988), who nds elasticities as low as 0.1 and often insignicantly dierent from zero. Using cohort
data, Attanasio and Weber (1993) obtain values for the elasticity of intertemporal substitution in the range 0.70.8,
while Beaudry and van Wincoop (1996) nd an elasticity close to one using a panel of data from U.S. states. Contrary
to these larger estimates, the survey evidence of Barsky, Juster, Kimball and Shapiro (1997) produces an estimate
of 0.18. An earlier version of the model presented here (Sheedy, 2013) has separate parameters for risk aversion and
intertemporal substitution, but quantitatively, the intertemporal substitution paramater is found to matter little for
the results. For this reason, the calibration of here focuses on its implications for risk aversion.
35
A simplifying feature of the model is that borrowers have the same marginal propensity to consume from nancial
wealth as savers in the neighbourhood of the steady state.
41
is lowest for housing wealth and larger for nancial wealth. Given the focus on nancial wealth in
this paper, the baseline calibration assumes m 6%, which using [5.3] implies 0.993.
36
The
sensitivity analysis considers marginal propensities to consume from 4% to 8%.
37
The range of available evidence on the Frisch elasticity of labour supply is discussed by Hall
(2009), who concludes that a value of approximately 2/3 is reasonable. However, both real business
cycle and New Keynesian models have typically assumed Frisch elasticities signicantly larger than
this, often as high as 4 (see, King and Rebelo, 1999, Rotemberg and Woodford, 1997). The baseline
calibration adopted here uses a Frisch elasticity of 2, and the sensitivity analysis considers a range
of values for from completely inelastic labour supply up to 4.
38
The debt maturity parameter (which stands in for the parameter specifying the sequence
of coupon payments, given = /(1 + n)) is set to match the average maturity of household debt
contracts. In the model, the average maturity of household debt is related to the duration of the
bond that is traded in incomplete nancial markets. Formally, duration T
m
refers to the average
of the maturities (in years) of each payment made by the bond weighted by its contribution to the
present value of the bond. Given the geometric sequence of nominal coupon payments parameterized
by , the bond duration (in steady state) is
T
m
=

=1
T

1
(1 +

j)

=
T
1

1+

j
,
using the steady-state bond price (present value of the coupon payments)

Q = (1 +

j)
1
from [3.12].
39
Let j denote the average annualized nominal interest rate on household debt, with
1 +

j = e
j T
. In the optimal policy analysis, the steady-state rate of ination will be zero ( = 0),
hence nominal GDP growth is n = g, and so = /(1 + g). It follows that and can be
determined by:
= e
j T
_
1
T
T
m
_
, and = e
gT
. [5.4]
Doepke and Schneider (2006) present evidence on the average duration of household nominal debt
liabilities. Their analysis takes account of renancing and prepayment of loans. For the most recent
36
Together with the baseline calibration of , , and , the original patience parameters are
b
1.006 and

s
1.012, and the implied value of is 1.009. Thus, the exogenous dierence between the annual rates of time
preference of borrowers and savers is approximately 2.4%.
37
A potential alternative approach to calibrating is to use its implications for the persistence of shocks to the
wealth distribution. In the model, the impulse response of the debt-to-GDP gap is proportional to

after time
periods have elapsed. The expected duration (in years) of the eects of shocks on the wealth distribution is thus
T
d
= T

=1
(1)
1
= T/(1), which can be used to obtain given an estimate of T
d
. The baseline calibration
is equivalent to T
d
36 years. The sensitivity analysis for the marginal propensity to consume implies a range of
values from 0.988 to 0.998, which is equivalent to considering values of T
d
from approximately 21 to 139 years.
38
Dierent Frisch elasticities for borrowers and savers are assumed to ensure that the wealth distribution has
no impact on the aggregate supply of labour. The required household-specic Frisch elasticities are
b
1.6 for
borrowers and
s
2.6 for savers.
39
Duration is equal to the percentage reduction in the real value of a nominal asset following a one percentage point
(annualized) permanent rise in ination. Since = /(1 + n),

j =

i, 1 +

i = (1 + )(1 + g), 1 + n = (1 + )(1 + g),

and = (1 + g)/(1 + ), it follows that T
m
= T/(1 ). A permanent rise in ination by one percentage point at
an annualized rate is equivalent to increasing
t
by T in all time periods from some date onwards, and this reduces
the ex-post real return on nominal bonds by T/(1 ), conrming the interpretation of T
m
.
42
year in their data (2004), the duration lies between 5 and 6 years, while the duration has not been
less than 4 years over the entire period covered by the study (19522004). This suggests a baseline
duration of T
m
5 years, which using [5.4] implies 0.967.
40
The sensitivity analysis considers
the eects of having durations as short as one quarter (one-period debt), and longer durations up
to 10 years.
41
There are two main strategies for calibrating the price elasticity of demand . The direct ap-
proach draws on studies estimating consumer responses to price dierences within narrow consump-
tion categories. A price elasticity of approximately three is typical of estimates at the retail level
(see, for example, Nevo, 2001), while estimates of consumer substitution across broad consump-
tion categories suggest much lower price elasticities, typically lower than one (Blundell, Pashardes
and Weber, 1993). Indirect approaches estimate the price elasticity based on the implied markup
1/( 1), or as part of the estimation of a DSGE model. Rotemberg and Woodford (1997) esti-
mate an elasticity of approximately 7.9 and point out this is consistent with the markups in the
range of 10%20%. Since it is the price elasticity of demand that directly matters for the welfare
consequences of ination rather than its implications for markups as such, the direct approach is
preferred here and the baseline value of is set to 3. A range of values from the theoretical minimum
elasticity of 1 up to 10 is considered in the sensitivity analysis.
The production function is given in equation [3.6]. If e denotes the elasticity of aggregate output
with respect to hours then the elasticity of real marginal cost with respect to output can be
obtained from e using:
=
1 e
e
.
A conventional value of e 2/3 is adopted for the baseline calibration (this would be the labour
share in a model with perfect competition), which implies 0.5. As discussed in Rotemberg
and Woodford (1999), there may be reasons to expect an elasticity of marginal cost with respect
to output higher than this (for example, if the elasticity of substitution between labour and other
factors is less than one), so the sensitivity analysis examines the eects of higher values of . An
important implication of is the strength of real rigidities (related to the term 1 + appearing
in the formula for in the Phillips curve [4.6a]), which are absent in the special case of a linear
production function ( = 0).
42
The sensitivity analysis considers values of between 0 and 1.
In the model, is the probability of not changing price in a given time period. The probability
40
A conventional T
f
-year xed-rate mortgage has a duration of (e
r T
f
1r T
f
)/(r (e
r T
f
1)), which is approximately
11 years with T
f
= 30 and r 5%. The calibrated duration may seem short given the high share of mortgage debt
in total household debt and the prevalence of 30-year xed-rate mortgages, but renancing shortens the duration of
debt. In the model, the frequency of renancing is determined by 1 . The baseline calibration implies that 12.6%
of the total stock of debt is renanced or newly issued each year.
41
The baseline value of is 0.971. The calibration method implicitly assumes is a structural parameter that
will remain invariant to changes in policy, including the change in the average rate of ination. An alternative is to
assume is the structural parameter, in which case is calibrated by dividing from [5.4] by 1 + n, where n is the
average of actual nominal GDP growth. This method leads to 0.958, which is well within the range of values of
considered in the sensitivity analysis.
42
It is conventional to assume a source of real rigidities in New Keynesian models, though Bils, Klenow and Malin
(2012) present some critical evidence.
43
distribution of survival times for newly set prices is (1 )

, and hence the expected duration of a

price spell T
p
(in years) is T
p
= T

=1
(1 )
1
= T/(1 ). With data on T
p
, the parameter
can be inferred from:
= 1
T
T
p
. [5.5]
There is now an extensive literature measuring the frequency of price adjustment across a represen-
tative sample of goods. Using the dataset underlying the U.S. CPI index, Nakamura and Steinsson
(2008) nd the median duration of a price spell is 79 months, excluding sales but including product
substitutions. Klenow and Malin (2010) survey a wide range of studies reporting median durations
in a range from 34 months to one year. The baseline duration is taken to be 8 months (T
p
8/12),
implying 0.625. The sensitivity analysis considers average durations from 3 to 15 months.
43
5.2 Results
Consider an economy hit by an unexpected permanent fall in potential output. How should monetary
policy react? In the basic New Keynesian model with sticky prices but either complete nancial
markets or a representative household, the optimal monetary policy response to a TFP shock is to
keep ination on target and allow actual output to fall in line with the loss of potential output.
Using the baseline calibration from Table 1 and the solution to the optimal monetary policy problem,
Figure 1 shows the impulse responses of the debt-to-GDP gap

d
t
, ination
t
, the output gap

Y
t
, and
the bond yield j
t
under the optimal monetary policy and under a policy of strict ination targeting
for the 30 years following a 10% fall in potential output.
With strict ination targeting, the debt-to-GDP gap rises in line with the fall in output (10%)
because the denominator of the debt-to-GDP ratio falls, while the numerator is unchanged. The
eects of this shock on the wealth distribution and hence on consumption are long lasting (the
half-life of the debt-to-GDP impulse response is around 25 years). Under strict ination targeting,
the output gap and bond yields are completely unchanged following the shock.
The optimal monetary policy response is in complete contrast to strict ination targeting. Op-
timal policy allows ination to rise, which stabilizes nominal GDP over time in spite of the fall in
real GDP. This helps to stabilize the debt-to-GDP ratio, moving the economy closer to the outcome
with complete nancial markets where borrowers would be insured against the shock and the value
of debt liabilities would automatically move in line with income. The rise in the debt-to-GDP gap
is very small (around 1%) compared to strict ination targeting (10%). The rise in ination is
very persistent, lasting around two decades. The higher ination called for is signicant, but not
43
An alternative approach to calibrating the parameters , , and related to nominal and real rigidities would be
to choose values consistent with estimates of the slope of the Phillips curve. The recent literature on estimating the
New Keynesian Phillips curve studies the relationship
t
= E
t

t+1
+(1/)x
t
between ination
t
and real marginal
cost x
t
, where the latter is proxied by the labour share. The baseline calibration implies 1/ 0.091. Gal and
Gertler (1999) present a range of estimates of 1/ lying between 0.02 and 0.04. Gal, Gertler and Lopez-Salido (2001)
estimate 1/ to be in the range 0.030.04, while Sbordone (2002) obtains an estimate of 0.055. The Phillips curve
implied by the baseline calibration is steeper than these estimates, but the sensitivity analysis for , , and does
allow for Phillips curve slopes in the range of econometric estimates. The maximum value of considered implies
1/ 0.021, the maximum value of implies 1/ 0.038, and the maximum value of implies 1/ 0.057.
44
Figure 1: Responses to a TFP shock, optimal monetary policy and strict ination targeting
Debt-to-GDP gap
%
Inflation
%
Ination target
Optimal policy
Output gap
Years
%
Bond yield
Years
%
0 10 20 30 0 10 20 30
0 10 20 30 0 10 20 30
0
0.5
1
0
0.05
0.1
0.15
0.2
0
1
2
3
0
2
4
6
8
10
Notes: The shock is an unexpected permanent TFP shock that reduces the natural level of output by 10%
relative to its trend. The debt-to-GDP gap and the output gap are reported as percentage deviations,
and ination and bond yields are reported as annualized percentage rates. The parameters are set in
accordance with the baseline calibration from Table 1.
dramatic: for the rst two years, around 23% higher, for the next decade around 12% higher,
and for the decade after that, around 01% higher. Ination that is spread out over time is still
eective in reducing the debt-to-GDP ratio because debt liabilities have a long average maturity. It
is also signicantly less costly in terms of relative-price distortions to have ination spread out over
a longer time than the typical durations of stickiness of individual prices.
The rise in ination does lead to a disturbance to the output gap for the rst one or two years,
but this is short lived because the duration of the real eects of monetary policy through the
traditional price-stickiness channel is brief compared to the relevant time scale of decades for the
other variables. The eect is also quantitatively small because ination is highly persistent, the
rise in expected ination closely matching the rise in actual ination, so the Phillips curve implies
little impact on the output gap. Finally, nominal bond yields show a persistent increase. It might
seem surprising that yields do not fall as monetary policy is loosened, but the bonds in question are
45
long-term bonds, and the eect on ination expectations is dominant (there is a fall in real interest
rates because the rise in bond yields is less than what would be implied by higher expected ination
alone).
The impulse response function of the debt-to-GDP gap under the optimal policy is proportional
to the impulse response under strict ination targeting. Since the debt-to-GDP gap would be zero
if incomplete nancial markets were the only concern of the policymaker, this provides a measure of
the weights attached by optimal policy to stabilizing the debt gap and to stabilizing ination. The
response of the debt gap under the optimal policy is approximately 11.6% of the response under
strict ination targeting, so the policy weight on debt gap stabilization is 88.4% and the policy
weight 1 on ination stabilization is 11.6% (similarly, the ination response under the optimal
policy is around 88.4% of what would keep the debt gap exactly at zero).
The baseline calibration implies that addressing the problem of incomplete nancial markets
is quantitatively the main focus of optimal monetary policy rather than other objectives such as
ination stabilization. What explains this, and how sensitive is this conclusion to the particular
calibration targets? Consider the exercise of varying each calibration target individually over the
ranges discussed in section 5.1, holding all other targets constant. For each new target, the implied
parameters are recalculated and the new policy weight is obtained. Figure 2 plots the values of
(the optimal policy weight on the debt-to-GDP gap) obtained for each target.
As can be seen in Figure 2, over the range of reasonable average real GDP growth rates there is
almost no eect on the optimal policy weight. The range of real interest rates is somewhat larger
(because there is less certainty about the appropriate real interest rate to assume for household
borrowing) but the optimal policy weight on incomplete nancial markets changes little. Both
average real growth and average real interest rates aect the discount factor , which enters the
equations of the model in many places, but there is no intuitively obvious reason to expect it to
have a large impact on the relative benets and costs of achieving the various objectives of policy.
The results are most sensitive to the calibration targets for the average debt-to-GDP ratio and
the coecient of relative risk aversion. The average debt-to-GDP ratio proxies for the parameter ,
which is related to the dierence in patience between borrowers and savers. It is not surprising that
an economy with less debt in relation to income has less of a concern with the incompleteness of
nancial markets because it means the impact of shocks is felt more evenly by borrowers and savers.
In the limiting case of a representative-household economy, the average debt-to-GDP ratio tends to
zero, and the degree of completeness of nancial markets becomes irrelevant. The debt gap receives
more than half the weight in the optimal policy as long as the calibration target for the average
debt-to-GDP ratio is not below 50%. It seems unlikely the U.S. would return to such low levels of
household debt in the foreseeable future if the levels of borrowing experienced since the 1990s do
indeed reect the preferences and income proles of borrowers and savers.
It is also not surprising that the results are sensitive to the coecient of relative risk aversion.
Since the only use for complete nancial markets in the model is risk sharing, if households were
risk neutral then there would be no loss from these markets being absent, as long as saving and
borrowing incomplete nancial markets remained possible. The baseline coecient of relative risk
46
Figure 2: Sensitivity analysis for optimal policy weight on incomplete nancial markets
Real GDP growth rate (g, annual %) Real interest rate (r , annual %)
Debt-to-GDP ratio (D, % of annual GDP) Coecient of relative risk aversion ()
Marginal propensity to consume (m, annual %) Frisch elasticity of labour supply ()
Duration of debt (T
m
, years) Price elasticity of demand ()
Marginal cost elasticity w.r.t. output () Duration of price stickiness (T
p
, years)
0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8
2 4 6 8 10 2 4 6 8 10
0 1 2 3 4 4 5 6 7 8
0 2 4 6 8 0 50 100 150 200
4 5 6 7 0 0.5 1 1.5 2
0.25
0.5
0.75
1
0.25
0.5
0.75
1
0.25
0.5
0.75
1
0.25
0.5
0.75
1
0.25
0.5
0.75
1
0.25
0.5
0.75
1
0.25
0.5
0.75
1
0.25
0.5
0.75
1
0.25
0.5
0.75
1
0.25
0.5
0.75
1
Notes: The response of the debt gap under the optimal policy is 1 multiplied by its response under
strict ination targeting. Each of the calibration targets in Table 1 is varied individually, holding all
others at their baseline values. The baseline value of is 0.884.
aversion is higher than the typical value of 2 found in many macroeconomic models (though that
number is usually relevant for intertemporal substitution in those models, not for attitudes to risk),
but it is low compared to the values often assumed in nance models that seek to match risk premia
47
(even the maximum value of 10 considered here would be insucient to generate realistic risk premia
without adding other features to the model). The optimal policy weight on the debt gap exceeds
one half if the coecient of relative risk aversion exceeds 1.3, so lower degrees of risk aversion do
not necessarily overturn the conclusions of this paper.
The next most important calibration target is the price elasticity of demand. A higher price
elasticity increases the welfare costs of ination. Welfare ultimately depends on quantities not
prices, but the price elasticity determines how much quantities are distorted by dispersion of relative
prices. To reduce the optimal policy weight on the debt gap below one half it is necessary to
assume price elasticities in excess of 10. Such values would be outside the range typical in IO and
microeconomic studies of demand, with 10 itself being at the high end of the range of values used
in most macroeconomic models. The typical value of 6 often found in New Keynesian models only
reduces to approximately 71%.
The results are largely insensitive to the marginal propensity to consume from nancial wealth,
which is used to determine the parameter in the specication of the endogenous discount factors.
Since this feature of the model was introduced only for technical reasons, it is reassuring that it does
not have a signicant eect on the results within a wide range of reasonable parameter values. The
Frisch elasticity of labour supply has a fairly small but not insignicant eect on the results, with
the optimal policy weight on the debt gap increasing with the Frisch elasticity. A higher elasticity
increases the welfare costs of shocks to wealth distribution by distorting the labour supply decisions
of dierent households, as well as making it easier for monetary policy to inuence the real value of
debt by changing the ex-ante real interest rate in addition to ination. An elastic labour supply does
mean that ination uctuations lead to output gap uctuations, which increases the importance of
targeting ination, but the rst two eects turn out to be more important quantitatively.
The results are somewhat more sensitive to the average duration of a price spell and the elasticity
of real marginal cost with respect to output. The rst of these determines the importance of nominal
price rigidities. Greater nominal rigidity leads to more dispersion of relative prices from a given
amount of ination, and thus reduces the optimal policy weight on the debt gap. A higher output
elasticity of marginal cost implies that the production function has greater curvature, so a given
dispersion of output levels across rms represents a more inecient allocation of resources. However,
the range of reasonable values for the duration of price stickiness does not reduce below 65%, and
the range of marginal cost elasticities does not lead to any value below 80%.
Finally, there is the average duration of household debt, where the eects of this calibration
target are more subtle. It might be expected that the longer the maturity of household debt, the
higher is the optimal policy weight on the debt gap. This is because longer-term debt allows ination
to be spread out further over time, reducing the welfare costs of the ination, yet still having an
eect on the real value of debt. However, the sensitivity analysis shows the optimal policy weight is
a non-monotonic function of debt maturity: either very short-term or long-term debt maturities lead
to high values of , while debt of around 1.5 years maturity has the lowest value of (approximately
75%).
This puzzle is resolved by recalling there are two ways monetary policy can aect the real value
48
of debt: ination to change the ex-post real return on nominal debt, and changes in the ex-ante real
interest rate (nancial repression). As has been discussed, the rst method is eective at a lower
cost for long debt maturities. When labour supply is inelastic, the second method is not available,
and the value of is then indeed a strictly increasing function of debt maturity (with the value of
falling to 15% for the shortest-maturity debt).
When the ex-ante real interest rate method is available, it is most eective (taking account of the
costs of using it in terms of ination and output gap uctuations) when debt maturities are short.
This is because monetary policy can only aect ex-ante real interest rates for a few years at most
(in line with reasonable calibrations of nominal and real rigidities). If debt is continually renanced
each quarter or comprises oating-rate instruments, monetary policy has signicant power to aect
its real value because nominal rigidities allow it to change the current real interest rate. However,
if xed-rate debt is rarely renanced then, holding ination constant (that is, ignoring the rst
method for aecting the real value of the debt), the real return is largely predetermined and thus
insensitive to current monetary policy. Therefore, intermediate debt maturities correspond to the
lowest optimal policy weights on stabilizing the debt gap because the maturity is too short for the
ination method to be eective at low cost, but too long for the ex-ante real interest rate method
to work.
6 Conclusions
This paper has shown how a monetary policy of nominal GDP targeting facilitates ecient risk
sharing in incomplete nancial markets where contracts are denominated in terms of money. In an
environment where risk derives from uncertainty about future real GDP, strict ination targeting
would lead to a very uneven distribution of risk, with leveraged borrowers consumption highly
exposed to any unexpected change in their incomes when monetary policy prevents any adjustment
of the real value of their liabilities. Strict ination targeting does provide savers with a risk-free real
return, but fundamentally, the economy lacks any technology that delivers risk-free real returns, so
the safety of savers portfolios is simply the ip-side of borrowers leverage and high levels of risk.
Absent any changes in the physical investment technology available to the economy, aggregate risk
cannot be annihilated, only redistributed.
That leaves the question of whether the distribution of risk is ecient. The combination of
incomplete markets and strict ination targeting implies a particularly inecient distribution of
risk when households are risk averse. If complete nancial markets were available, borrowers would
issue state-contingent debt where the contractual repayment is lower in a recession and higher in
a boom. These securities would resemble equity shares in GDP, and they would have the eect
of reducing the leverage of borrowers and hence distributing risk more evenly. In the absence of
such nancial markets, in particular because of the inability of households to sell such securities,
a monetary policy of nominal GDP targeting can eectively complete nancial markets even when
only non-contingent nominal debt is available. Nominal GDP targeting operates by stabilizing the
debt-to-GDP ratio. With nancial contracts specifying liabilities xed in terms of money, a policy
49
that stabilizes the monetary value of real incomes ensures that borrowers are not forced to bear too
much aggregate risk, converting nominal debt into real equity.
While the model is far too simple to apply to the recent nancial crises and deep recessions
experienced by a number of economies, one policy implication does resonate with the predicament of
several economies faced with high levels of debt combined with stagnant or falling GDPs. Nominal
GDP targeting is equivalent to a countercyclical price level, so the model suggests that higher
ination can be optimal in recessions. In other words, while each of the stagnation and ination
that make up the word stagation is bad in itself, if stagnation cannot immediately be remedied,
some ination might be a good idea to compensate for the ineciency of incomplete nancial
markets. And even if policymakers were reluctant to abandon ination targeting, the model does
suggest that they have the strongest incentives to avoid deation during recessions (a procyclical
price level). Deation would raise the real value of debt, which combined with falling real incomes
would be the very opposite of the risk sharing stressed in this paper, and even worse than an
unchanging ination rate.
It is important to stress that the policy implications of the model in recessions are matched by
equal and opposite prescriptions during an expansion. Thus, it is not just that optimal monetary
policy tolerates higher ination in a recession it also requires lower ination or even deation
during a period of high growth. Pursuing higher ination in recessions without following a symmetric
policy during an expansion is both inecient and jeopardizes an environment of low ination on
average. Therefore the model also argues that more should be done by central banks to take away
the punch bowl during a boom even were ination to be stable.
References
Abel, A. B. (1990), Asset prices under habit formation and catching up with the Joneses,
American Economic Review, 80(2):3842 (May). 25
Akyol, A. (2004), Optimal monetary policy in an economy with incomplete markets and idiosyn-
cratic risk, Journal of Monetary Economics, 51(6):12451269 (September). 6
Allen, F., Carletti, E. and Gale, D. (2011), Money, nancial stability and eciency,
Working paper, New York University. 6
Andr

es, J., Arce, O. J. and Thomas, C. (2010), Banking competition, collateral constraints
and optimal monetary policy, Working paper 1001, Bank of Spain. 6
Attanasio, O. and Davis, S. J. (1996), Relative wage movements and the distribution of
consumption, Journal of Political Economy, 104(6):12271262 (December). 22
Attanasio, O. P. and Weber, G. (1993), Consumption growth, the interest rate and aggrega-
tion, Review of Economic Studies, 60(3):631649 (July). 41
Bach, G. L. and Stephenson, J. B. (1974), Ination and the redistribution of wealth, Review
of Economics and Statistics, 56(1):113 (February). 5
50
Bailey, S. (1837), Money and Its Vicissitudes in Value, Engham Wilson, London. 24
Bansal, R. and Yaron, A. (2004), Risks for the long run: A potential resolution of asset pricing
puzzles, Journal of Finance, 59(4):14811509 (August). 41
Barro, R. J. (1979), On the determination of the public debt, Journal of Political Economy,
87(5):940971 (October). 16, 35
(2006), Rare disasters and asset markets in the twentieth century, Quarterly Journal of
Economics, 121(3):823866. 41
Barsky, R. B., Juster, T. F., Kimball, M. S. and Shapiro, M. D. (1997), Preference
parameters and behavioral heterogeneity: An experimental approach in the health and retirement
study, Quarterly Journal of Economics, 112(2):537579 (May). 23, 41
Bean, C. R. (1983), Targeting nominal income: An appraisal, Economic Journal, 93:806819
(December). 7
Beaudry, P. and van Wincoop, E. (1996), The intertemporal elasticity of substitution: An
exploration using a US panel of state data, Economica, 63:495512 (August). 41
Bernanke, B. and Gertler, M. (1989), Agency costs, net worth, and business uctuations,
American Economic Review, 79(1):1431 (March). 1
Bernanke, B. S. and Gertler, M. (2001), Should central banks respond to movements in
asset prices?, American Economic Review, 91(2):253257 (May). 3
Bernanke, B. S., Gertler, M. and Gilchrist, S. (1999), The nancial accelerator in a
quantitative business cycle framework, in J. B. Taylor and M. Woodford (eds.), Handbook of
Macroeconomics, vol. 1C, Elsevier, chap. 21, pp. 13411393. 1
Bernanke, B. S., Laubach, T., Mishkin, F. S. and Posen, A. S. (1999), Ination Targeting:
Lessons from the International Experience, Princeton University Press. 1
Bils, M., Klenow, P. J. and Malin, B. A. (2012), Reset price ination and the impact of
monetary policy shocks, American Economic Review, 102(6):27982825 (October). 43
Blundell, R., Pashardes, P. and Weber, G. (1993), What do we learn about consumer
demand patterns from micro data?, American Economic Review, 83(3):570597 (June). 43
Bohn, H. (1988), Why do we have nominal government debt?, Journal of Monetary Economics,
21(1):127140 (January). 6
Calvo, G. A. (1983), Staggered prices in a utility-maximizing framework, Journal of Monetary
Economics, 12(3):383398 (September). 27
Carlstrom, C. T., Fuerst, T. S. and Paustian, M. (2010), Optimal monetary policy in a
model with agency costs, Journal of Money, Credit and Banking, 42(s1):3770 (September). 1
Chari, V. V., Christiano, L. J. and Kehoe, P. J. (1991), Optimal scal and monetary policy:
Some recent results, Journal of Money, Credit and Banking, 23(3(2)):519539 (August). 6
Chari, V. V. and Kehoe, P. J. (1999), Optimal scal and monetary policy, in J. B. Taylor
and M. Woodford (eds.), Handbook of Monetary Economics, vol. 1C, Elsevier, chap. 26, pp.
16711745. 6
51
Christiano, L., Ilut, C. L., Motto, R. and Rostagno, M. (2010), Monetary policy and
stock market booms, Working paper 16402, NBER. 3
Christiano, L., Motto, R. and Rostagno, M. (2010), Financial factors in economic uctu-
ations, Working paper 1192, European Central Bank. 1
Clarida, R., Gal

, J. and Gertler, M. (1999), The science of monetary policy: A New

Keynesian perspective, Journal of Economic Literature, 37(4):16611707 (December). 22
Cochrane, J. H. (1991), A simple test of consumption insurance, Journal of Political Economy,
99(5):957976 (October). 22
Cohen, A. and Einav, L. (2007), Estimating risk preferences from deductible choice, American
Economic Review, 97(3):745788 (June). 23, 41
Cole, H. L. and Kocherlakota, N. R. (2001), Ecient allocations with hidden income and
hidden storage, Review of Economic Studies, 68(3):523542 (July). 21
Cooley, T. F. and Ohanian, L. E. (1991), The cyclical behavior of prices, Journal of Mone-
tary Economics, 28(1):2560 (August). 24
Cukierman, A., Lennan, K. and Papadia, F. (1985), Ination-induced redistributions via
monetary assets in ve European countries: 19741982, in J. Mortensen (ed.), Economic Policy
and National Accounting in Inationary Conditions, vol. 2 of Studies in Banking and Finance,
North-Holland, Amsterdam. 5
C

urdia, V. and Woodford, M. (2009), Credit frictions and optimal monetary policy, Working
paper, Columbia University. 7, 39
De Fiore, F., Teles, P. and Tristani, O. (2011), Monetary policy and the nancing of
rms, American Economic Journal: Macroeconomics, 3(4):112142 (October). 6
De Fiore, F. and Tristani, O. (2012), Optimal monetary policy in a model of the credit
channel, Economic Journal, forthcoming. 1
Doepke, M. and Schneider, M. (2006), Ination and the redistribution of nominal wealth,
Journal of Political Economy, 114(6):10691097 (December). 5, 22, 42
Eggertsson, G. B. and Krugman, P. (2012), Debt, deleveraging, and the liquidity trap: A
Fisher-Minsky-Koo approach, Quarterly Journal of Economics, 127(3):14691513. 7
Erceg, C. J., Henderson, D. W. and Levin, A. T. (2000), Optimal monetary policy with
staggered wage and price contracts, Journal of Monetary Economics, 46(2):281313 (October).
27, 34
Faia, E. and Monacelli, T. (2007), Optimal interest rate rules, asset prices, and credit fric-
tions, Journal of Economic Dynamics and Control, 31(10):32283254 (October). 1
Gal

, J. and Gertler, M. (1999), Ination dynamics: a structural econometric analysis,

Journal of Monetary Economics, 44(2):195222 (October). 44
Gal

, J., Gertler, M. and L

opez-Salido, J. D. (2001), European ination dynamics,

European Economic Review, 45(7):12371270 (June). 44
52
Garriga, C., Kydland, F. E. and

Sustek, R. (2013), Mortgages and monetary policy,
Discussion paper 2013-6, Centre for Macroeconomics. 26
Ghent, A. C. (2011), Securitization and mortgage renegotiation: Evidence from the Great De-
pression, Review of Financial Studies, 24(6):18141847. 22
Guerrieri, V. and Lorenzoni, G. (2011), Credit crises, precautionary saving, and the liquidity
trap, Working paper 17583, NBER. 7
Hall, R. E. (1988), Intertemporal substitution in consumption, Journal of Political Economy,
96(2):339357 (April). 41
(2009), Reconciling cyclical movements in the marginal value of time and the marginal
product of labor, Journal of Political Economy, 117(2):281323 (April). 42
Hall, R. E. and Mankiw, N. G. (1994), Nominal income targeting, in N. G. Mankiw (ed.),
Monetary Policy, University of Chicago Press, chap. 2, pp. 7194. 7
Hansen, L. P. and Singleton, K. J. (1982), Generalized instrumental variables estimation of
nonlinear rational expectations models, Econometrica, 50(5):12691286 (September). 41
Hayashi, F., Altonji, J. and Kotlikoff, L. (1996), Risk-sharing between and within fami-
lies, Econometrica, 64(2):261294 (March). 22
Hoelle, M. and Peiris, M. U. (2013), On the eciency of nominal GDP targeting in a large
open economy, Working paper 1273, Krannert School of Management, Purdue University. 6
Iacoviello, M. (2005), House prices, borrowing constraints, and monetary policy in the business
cycle, American Economic Review, 95(3):739764 (June). 7
Juster, F. T., Lupton, J. P., Smith, J. P. and Stafford, F. (2006), The decline in house-
hold saving and the wealth eect, Review of Economics and Statistics, 88(1):2027 (February).
41
Kehoe, T. J. and Levine, D. K. (1993), Debt-constrained asset markets, Review of Economic
Studies, 60(4):865888 (October). 21
King, R. G. and Rebelo, S. T. (1999), Resuscitating real business cycles, in J. B. Taylor and
M. Woodford (eds.), Handbook of Macroeconomics, vol. 1B, Elsevier, chap. 14, pp. 9271007. 39,
42
Kiyotaki, N. and Moore, J. (1997), Credit cycles, Journal of Political Economy, 105(2):211
248 (April). 1
Klenow, P. J. and Malin, B. A. (2010), Microeconomic evidence on price-setting, in B. M.
Friedman and M. Woodford (eds.), Handbook of Monetary Economics, vol. 3, Elsevier, chap. 6,
pp. 231284. 44
Koenig, E. F. (2013), Like a good neighbor: Monetary policy, nancial stability, and the distri-
bution of risk, International Journal of Central Banking, 9(2):5782 (June). 6
Kryvtsov, O., Shukayev, M. and Ueberfeldt, A. (2011), Optimal monetary policy under
incomplete markets and aggregate uncertainty: A long-run perspective, Journal of Economic
Dynamics and Control, 35(7):10451060 (July). 6
53
Lee, J. W. (2010), Monetary policy with heterogeneous households and imperfect risk-sharing,
Working paper, Rutgers University. 6
Lucas, R. E., Jr. and Stokey, N. L. (1983), Optimal scal and monetary policy in an economy
without capital, Journal of Monetary Economics, 12(1):5593. 16
Lustig, H., Sleet, C. and Yeltekin, S. (2008), Fiscal hedging with nominal assets, Journal
of Monetary Economics, 55(4):710727 (May). 7
Meade, J. E. (1978), The meaning of internal balance, Economic Journal, 88:423435
(September). 7
Mehra, R. and Prescott, E. C. (1985), The equity premium: A puzzle, Journal of Monetary
Economics, 15(2):145161 (March). 40
Nakamura, E. and Steinsson, J. (2008), Five facts about prices: A reevaluation of menu cost
models, Quarterly Journal of Economics, 123(4):14151464 (November). 44
Nelson, J. A. (1994), On testing for full insurance using consumer expenditure survey data,
Journal of Political Economy, 102(2):384394 (April). 22
Nevo, A. (2001), Measuring market power in the ready-to-eat cereal industry, Econometrica,
69(2):307342 (March). 43
Osorio-Rodr

guez, D. (2013), Foreign currency debt and optimal monetary policy: Is there a
role for the exchange rate in completing markets?, Working paper, London School of Economics.
6
Parker, J. A. (1999), Spendthrift in America? On two decades of decline in the U.S. saving
rate, NBER Macroeconomics Annual, 14:317370. 41
Persson, T. and Svensson, L. E. O. (1989), Exchange rate variability and asset trade,
Journal of Monetary Economics, 23(3):485509 (May). 6
Pescatori, A. (2007), Incomplete markets and households exposure to interest rate and ination
risk: Implications for the monetary policy maker, Working paper 07-09, Federal Reserve Bank
of Cleveland. 6
Piskorski, T., Seru, A. and Vig, V. (2010), Securitization and distressed loan renegotiation:
Evidence from the subprime mortgage crisis, Journal of Financial Economics, 97(3):369397
(September). 22
Poterba, J. M. (2000), Stock market wealth and consumption, Journal of Economic Perspec-
tives, 14(2):99118 (Spring). 41
Rotemberg, J. J. and Woodford, M. (1997), An optimization-based econometric framework
for the evaluation of monetary policy, NBER Macroeconomics Annual, 12:297346. 42, 43
(1999), The cyclical behavior of prices and costs, in J. B. Taylor and M. Woodford (eds.),
Handbook of Macroeconomics, vol. 1B, Elsevier, chap. 16, pp. 10511135. 43
Sbordone, A. M. (2002), Prices and unit labor costs: a new test of price stickiness, Journal of
Monetary Economics, 49(2):265292 (March). 44
54
Schmitt-Groh

e, S. and Uribe, M. (2003), Closing small open economy models, Journal of

International Economics, 61(1):163185 (October). 25
(2004), Optimal scal and monetary policy under sticky prices, Journal of Economic
Theory, 114(2):198230 (February). 6
Selgin, G. (1995), The productivity norm versus zero ination in the history of economic
thought, History of Political Economy, 27(4):705735 (Winter). 6, 24
(1997), Less than zero: The case for a falling price level in a growing economy, Hobart
paper 132, Institute of Economic Aairs. 5
Sheedy, K. D. (2013), Debt and incomplete nancial markets: A case for nominal GDP target-
ing, Discussion paper 1209, Centre for Economic Performance. 25, 41
Shiller, R. J. (1993), Macro Markets: Creating Institutions for Managing Societys Largest Eco-
nomic Risks, Oxford University Press, New York. 21
(1997), Public resistance to indexation: A puzzle, Brookings Papers on Economic Activity,
1997(1):159228. 22, 24
Siu, H. E. (2004), Optimal scal and monetary policy with sticky prices, Journal of Monetary
Economics, 51(3):575607 (April). 6
Sumner, S. (2012), The case for nominal GDP targeting, Working paper, Mercatus Center,
George Mason University. 7
Uzawa, H. (1968), Time preference, the consumption function, and optimum asset holdings,
in J. N. Wolfe (ed.), Value, Capital and Growth: Papers in Honour of Sir John Hicks, Aldine,
Chicago. 25
Vlieghe, G. W. (2010), Imperfect credit markets: implications for monetary policy, Working
paper 385, Bank of England. 6
White, M. J. (2009a), Bankruptcy: Past puzzles, recent reforms, and the mortgage crisis,
American Law and Economics Review, 11(1):123. 22
White, W. R. (2009b), Should monetary policy lean or clean?, Working paper 34, Federal
Reserve Bank of Dallas, Globalization and Monetary Policy Institute. 3
Woodford, M. (2001), Fiscal requirements for price stability, Journal of Money, Credit and
Banking, 33(3):669728 (August). 26
(2003), Interest and Prices: Foundations of a Theory of Monetary Policy, Princeton Univer-
sity Press, New Jersey. 1, 27, 32
(2011), Ination targeting and nancial stability, Working paper, Columbia University. 3
A Appendices
Appendices are available in the full version of the paper:
http://personal.lse.ac.uk/sheedy/papers/DebtIncompleteFinancialMarketsAppendix.pdf
55